Infectious Diseases in Critical Care Medicine, Second Edition (Infectious Disease and Therapy)

  • 58 128 4
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up

Infectious Diseases in Critical Care Medicine, Second Edition (Infectious Disease and Therapy)

Infectious Diseases in Critical Care Medicine DK617x_C000a.indd 1 08/11/2006 8:10:09 AM INFECTIOUS DISEASE AND THERA

2,538 699 4MB

Pages 690 Page size 415.44 x 653.76 pts Year 2006

Report DMCA / Copyright

DOWNLOAD FILE

Recommend Papers

File loading please wait...
Citation preview

Infectious Diseases in Critical Care Medicine

DK617x_C000a.indd 1

08/11/2006 8:10:09 AM

INFECTIOUS DISEASE AND THERAPY Series Editor

Burke A. Cunha Winthrop-University Hospital Mineola, and State University of New York School of Medicine Stony Brook, New York

1. Parasitic Infections in the Compromised Host, edited by Peter D. Walzer and Robert M. Genta 2. Nucleic Acid and Monoclonal Antibody Probes: Applications in Diagnostic Methodology, edited by Bala Swaminathan and Gyan Prakash 3. Opportunistic Infections in Patients with the Acquired Immunodeficiency Syndrome, edited by Gifford Leoung and John Mills 4. Acyclovir Therapy for Herpesvirus Infections, edited by David A. Baker 5. The New Generation of Quinolones, edited by Clifford Siporin, Carl L. Heifetz, and John M. Domagala 6. Methicillin-Resistant Staphylococcus aureus: Clinical Management and Laboratory Aspects, edited by Mary T. Cafferkey 7. Hepatitis B Vaccines in Clinical Practice, edited by Ronald W. Ellis 8. The New Macrolides, Azalides, and Streptogramins: Pharmacology and Clinical Applications, edited by Harold C. Neu, Lowell S. Young, and Stephen H. Zinner 9. Antimicrobial Therapy in the Elderly Patient, edited by Thomas T. Yoshikawa and Dean C. Norman 10. Viral Infections of the Gastrointestinal Tract: Second Edition, Revised and Expanded, edited by Albert Z. Kapikian 11. Development and Clinical Uses of Haemophilus b Conjugate Vaccines, edited by Ronald W. Ellis and Dan M. Granoff 12. Pseudomonas aeruginosa Infections and Treatment, edited by Aldona L. Baltch and Raymond P. Smith 13. Herpesvirus Infections, edited by Ronald Glaser and James F. Jones 14. Chronic Fatigue Syndrome, edited by Stephen E. Straus 15. Immunotherapy of Infections, edited by K. Noel Masihi 16. Diagnosis and Management of Bone Infections, edited by Luis E. Jauregui 17. Drug Transport in Antimicrobial and Anticancer Chemotherapy, edited by Nafsika H. Georgopapadakou 18. New Macrolides, Azalides, and Streptogramins in Clinical Practice, edited by Harold C. Neu, Lowell S. Young, Stephen H. Zinner, and Jacques F. Acar

DK617x_C000a.indd 2

08/11/2006 8:10:09 AM

19. Novel Therapeutic Strategies in the Treatment of Sepsis, edited by David C. Morrison and John L. Ryan 20. Catheter-Related Infections, edited by Harald Seifert, Bernd Jansen, and Barry M. Farr 21. Expanding Indications for the New Macrolides, Azalides, and Streptogramins, edited by Stephen H. Zinner, Lowell S. Young, Jacques F. Acar, and Harold C. Neu 22. Infectious Diseases in Critical Care Medicine, edited by Burke A. Cunha 23. New Considerations for Macrolides, Azalides, Streptogramins, and Ketolides, edited by Stephen H. Zinner, Lowell S. Young, Jacques F. Acar, and Carmen Ortiz-Neu 24. Tickborne Infectious Diseases: Diagnosis and Management, edited by Burke A. Cunha 25. Protease Inhibitors in AIDS Therapy, edited by Richard C. Ogden and Charles W. Flexner 26. Laboratory Diagnosis of Bacterial Infections, edited by Nevio Cimolai 27. Chemokine Receptors and AIDS, edited by Thomas R. O’Brien 28. Antimicrobial Pharmacodynamics in Theory and Clinical Practice, edited by Charles H. Nightingale, Takeo Murakawa, and Paul G. Ambrose 29. Pediatric Anaerobic Infections: Diagnosis and Management, Third Edition, Revised and Expanded, Itzhak Brook 30. Viral Infections and Treatment, edited by Helga Ruebsamen-Waigmann, Karl Deres, Guy Hewlett, and Reinhold Welker 31. Community-Aquired Respiratory Infections, edited by Charles H. Nightingale, Paul G. Ambrose, and Thomas M. File 32. Catheter-Related Infections: Second Edition, Harald Seifert, Bernd Jansen and Barry Farr 33. Antibiotic Optimization: Concepts and Strategies in Clinical Practice (PBK), edited by Robert C. Owens, Jr., Charles H. Nightingale and Paul G. Ambrose 34. Fungal Infections in the Immunocompromised Patient, edited by John R. Wingard and Elias J. Anaissie 35. Sinusitis: From Microbiology To Management, edited by Itzhak Brook 36. Herpes Simplex Viruses, edited by Marie Studahl, Paola Cinque and Tomas Bergström 37. Antiviral Agents, Vaccines, and Immunotherapies, Stephen K. Tyring 38. Epstein-Barr Virus, Alex Tselis and Hal B. Jenson 39. Infection Management for Geriatrics in Long-Term Care Facilities, Second Edition, edited by Thomas T. Yoshikawa and Joseph G. Ouslander 40. Infectious Diseases in Critical Care Medicine, Second Edition, edited by Burke A. Cunha

DK617x_C000a.indd 3

08/11/2006 8:10:09 AM

DK617x_C000a.indd 4

08/11/2006 8:10:09 AM

Infectious Diseases in Critical Care Medicine Second Edition

edited by

Burke A. Cunha

Winthrop-University Hospital Mineola and State University of New York School of Medicine Stony Brook, New York, U.S.A.

New York London

DK617x_C000a.indd 5

08/11/2006 8:10:09 AM

Informa Healthcare USA, Inc. 270 Madison Avenue New York, NY 10016 © 2007 by Informa Healthcare USA, Inc. Informa Healthcare is an Informa business No claim to original U.S. Government works Printed in the United States of America on acid‑free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number‑10: 0‑8493‑3617‑1 (Hardcover) International Standard Book Number‑13: 978‑0‑8493‑3617‑1 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any informa‑ tion storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http:// www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978‑750‑8400. CCC is a not‑for‑profit organization that provides licenses and registration for a variety of users. For orga‑ nizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging‑in‑Publication Data Infectious diseases in critical care medicine / edited by Burke A. Cunha. ‑‑2nd ed. p. ; cm. ‑‑ (Infectious disease and therapy ; 40) Includes bibliographical references and index. ISBN‑13: 978‑0‑8493‑3617‑1 (hardcover : alk. paper) ISBN‑10: 0‑8493‑3617‑1 (hardcover : alk. paper) 1. Nosocomial infections. 2. Critical care medicine. 3. Intensive care units. I. Cunha, Burke A. II. Series. [DNLM: 1. Communicable Diseases‑‑diagnosis. 2. Critical Care. 3. Diagnosis, Differential. 4. Intensive Care Units. W1 IN406HMN v.40 2006 / WC 100 I4165 2006] RC112.I4595 2006 616.9’0475‑‑dc22

2006046567

Visit the Informa Web site at www.informa.com and the Informa Healthcare Web site at www.informahealthcare.com

DK617x_C000a.indd 6

08/11/2006 8:10:10 AM

For Marie Peerless wife and mother, Provider of domestic peace and harmony, Paragon of truth and beauty, Paradigm of earthly perfection . . . With gratitude for love and constant support.

Foreword

In the United States, there are over 4000 intensive care units containing 87,000 beds. While the number of hospitals has been more than decreasing in the United States over the past decade, the number of intensive care unit beds has increased. From 1985 to 2000, while the total number of U.S. hospitals decreased by 8.9% and total beds decreased by 26.4%, the number of intensive care unit beds increased by 26.2% (1–3). Intensive care unit beds have a high occupancy rate (over 65% nationally). The acuity of patients in those beds is rising. The costs associated with intensive care unit care have risen at a faster rate from 1985 to 2000 (190%) than the cost of hospital care in general (150%) and the gross domestic product (133%). Analyses have estimated that intensive care unit care cost the United States $33 billion to $55 billion in 1995 (3–5). Thus, these numbers give a numerical background to the trend obvious to all health care providers: a larger and larger fraction of hospitalized patients have high acuity and require extensive resources if the patients needing such care are to benefit from the impressive advances medical science has made over the past several decades. Intensive care units have become such an integral part of the health care complex over the past 40 years because medical science has developed the technical and cognitive skills to reverse life-threatening processes with an increasingly higher success rate than was conceivable even one or two decades ago. Victims of trauma, cancer, acute infections, myocardial damage, hemorrhagic diatheses, renal failure, hepatic failure, or neurologic catastrophes are examples of patients who have a far better prognosis in 2005 if they can have prompt access to modern critical care facilities. In terms of the infections that bring patients to intensive care units, patient outcome has improved substantially. These improvements include a more sophisticated understanding of pathophysiology, more accurate and less invasive diagnostic approaches, and a dramatic expansion of therapeutic options. These improvements have permitted patients to survive despite sepsis, human immunodeficiency virus infection, fungal disease, or viral processes to a degree not imagined two decades ago. Many other patients, however, are admitted to intensive care units with noninfectious catastrophes, only to have their courses complicated or their lives terminated by infections they acquired in the intensive care unit. Thus, ventilator-associated pneumonia, intravascular catheter-associated sepsis, urosepsis due to a urethral catheter or fungal superinfection, and candidemia are examples of processes that can be v

vi

Foreword

prevented to a substantial degree by improved strategies to prevent such complications. When infections do occur, they are more likely to be due to antibiotic-resistant pathogens because of heavy antibiotic exposure in the intensive care unit and the contamination of patients with pathogens transmitted from other sick patients by aerosols, droplets, fomites, or health care staff. As many as 45% of nosocomial infections occur in intensive care unit patients, although intensive care unit beds account for only 13% of all hospital beds. In this book edited by Burke A. Cunha, 32 chapters summarize the current knowledge concerning infectious diseases occurring in the field of critical care medicine. The material described in this book is written by an impressive team of experienced clinicians who deal regularly with patients in intensive care units. The material covered in these chapters is a compilation of information on general concepts, specific syndromes, and current management strategies that combines material that is derived from disparate sources. For intensivists, and for hospitalists, house officers, and consultants who spend much of their time caring for patients in intensive care units, this focused book will be most useful. Critical care medicine is an expanding aspect of health care in the developed world. Proper management of infectious complications is an important priority, given the magnitude and impact of the infections on patient outcome. Regular updates of this book will be most useful. Henry Masur Department of Critical Care National Institutes for Health Bethesda, Maryland, U.S.A.

REFERENCES 1. Angus DC, Kelley MA, Schmitz RJ, White A, Popovich J Jr., Committee von Manpower for Pulmonary and Critical Care Societies (COMPACCS). Caring for the critically ill patient. Current and projected workforce requirements for care of the critically ill and patients with pulmonary disease: can we meet the requirements of an aging population? JAMA 2000; 284(21):2762–2770. 2. Cooper RA. The COMPACCS Study: questions left unanswered. The Committee on Manpower for Pulmonary and Critical Care Societies. Am J Respir Crit Care Med 2001; 163(1):10–11. 3. Halpern NA, Pastores SM, Greenstein RJ. Critical care medicine in the United States 1985– 2000: an analysis of bed numbers, use, and costs. Crit Care Med 2004; 32:1254–1259. 4. Jacobs P, Noseworthy TW. National estimates of intensive care utilization and costs: Canada and the United States. Crit Care Med 1990; 18:1282–1286. 5. Halpern N, Bettes I, Greenstein R. Federal and nationwide intensive care units and health care costs: 1986–1992. Crit Care Med 1994; 4:2366–2373.

Preface

Infectious diseases continue to represent a major diagnostic and therapeutic challenge in the critical care unit. Infectious diseases maintain their preeminence in the critical care unit setting because of their frequency and importance in the critical care unit patient population. Since the first edition of Infectious Disease in Critical Care Medicine, there have been newly described infectious diseases to be considered in differential diagnosis, and new antimicrobial agents have been added to the therapeutic armamentarium. The second edition of Infectious Diseases in Critical Care Medicine continues the clinical orientation of the first edition. Differential diagnostic considerations in infectious diseases continue to be the central focus of the second edition. Clinicians caring for acutely ill patients in the critical care unit are confronted with the common problem of differentiating noninfectious disease mimics from their infectious disease counterparts. For this reason, the differential diagnosis of noninfectious diseases remains an important component of infectious diseases in the second edition. The second edition of Infectious Diseases in Critical Care Medicine emphasizes differential clinical features that enable clinicians to sort out complicated diagnostic problems. Because critical care unit patients often have complicated/interrelated multisystem disorders, subspecialty expertise is essential for optimal patient care. Early utilization of infectious disease consultation is important to assure proper application/ interpretation of appropriate laboratory tests and for the selection/optimization of antimicrobial therapy. Selecting the optimal antimicrobial for use in the critical care unit is vital. As important is the optimization of antimicrobial dosing to take into account the pharmacokinetic and pharmacodynamic attributes of the antibiotic. The infectious disease clinician, in addition to optimizing dosing considerations, is also able to evaluate potential antimicrobial side effects as well as drug–drug interactions, which may affect therapy. Infectious disease consultations can be helpful in differentiating colonization ordinarily not treated from infection that should be treated. Physicians who are not infectious disease clinicians lack the necessary sophistication in clinical infectious disease training, medical microbiology, pharmacokinetics/ pharmacodynamics, and diagnostic experience. Physicians in critical care units should rely on infectious disease clinicians as well as other consultants to optimize care for these acutely ill patients. vii

viii

Preface

The second edition of Infectious Diseases in Critical Care Medicine has been streamlined while maintaining its clinical focus. Again, the authors have been selected for their expertise and experience. The contributors are world-class teachers/ clinicians who have, in their writings, imparted their clinical experience for the benefit of the critical care unit physicians and their patients. The second edition of Infectious Diseases in Critical Care Medicine remains the only book dealing with infections in critical care. Burke A. Cunha

Contents

Foreword Henry Masur . . . . v Preface . . . . vii Contributors . . . . xvii PART I: GENERAL CONCEPTS 1. Methicillin-Resistant Staphylococcus aureus/Vancomycin-Resistant Enterococci Colonization and Infection in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 C. Glen Mayhall Introduction . . . . 1 Methicillin-Resistant S. aureus . . . . 2 Vancomycin-Resistant Enterococci . . . . 12 References . . . . 23 2. Outbreak Investigation in the Critical Care Unit . . . . . . . . . . . . . . 33 Brian W. Cooper Introduction . . . . 33 Surveillance . . . . 34 Investigation of Clusters of Infection . . . . 34 Notification of Appropriate Individuals . . . . 35 Construct an Epidemic Curve . . . . 36 Review the Literature . . . . 36 Develop a Line Listing . . . . 36 Develop Hypotheses and Institute Preliminary Controls . . . . 37 Evaluate the Effectiveness of Control Measures . . . . 37 Further Studies . . . . 38 Summary . . . . 38 References . . . . 38 3. Clinical Approach to Fever in the Critical Care Unit . . . . . . . . . . . 41 Burke A. Cunha Introduction . . . . 41 Diagnostic Considerations . . . . 42 Causes of Fever in the CCU . . . . 42 ix

x

Contents

Clinical Significance of Fever Patterns . . . . 45 Pulse–Temperature Relationships . . . . 50 Diagnostic Significance of Fever Defervescence Patterns . . . . 54 Obscure Fevers in the CCUs . . . . 56 Clinical Diagnostic Approach to Fever in the CCU . . . . 60 Clinical Therapeutic Approach . . . . 64 References . . . . 70 4. Sepsis and Its Mimics in the Critical Care Unit . . . . . . . . . . . . . . 73 Burke A. Cunha Introduction . . . . 73 Laboratory Abnormalities in Sepsis . . . . 74 Empiric Antimicrobial Therapy . . . . 77 Summary . . . . 78 References . . . . 78 PART II: CLINICAL SYNDROMES 5. Meningitis and Its Mimics in the Critical Care Unit . . . . . . . . . . . 81 Burke A. Cunha Overview . . . . 81 Clinical Diagnosis of Acute Bacterial Meningitis . . . . 82 The Mimics of Meningitis . . . . 83 Clinical and Laboratory Features of Acute Bacterial Meningitis . . . . 86 Empiric Therapy of Acute Bacterial Meningitis . . . . 98 References . . . . 102 6. Encephalitis and Its Mimics in the Critical Care Unit . . . . . . . . . 105 Burke A. Cunha Introduction . . . . 105 Mimics of Encephalitis in the CCU . . . . 106 Noninfectious Mimics of Acute Viral Encephalitis . . . . 109 Miscellaneous Other Disorders . . . . 110 Viral Causes of Acute Encephalitis in the CCU . . . . 113 Clinical Diagnostic Approach . . . . 113 Nonviral Infectious Mimics of Acute Encephalitis in the CCU . . . . 117 Treatable Causes of Acute Encephalitis in the CCU . . . . 118 Acute Encephalitis in Normal Hosts . . . . 118 Acute Encephalitis in Compromised Hosts . . . . 119 Therapeutic Approach . . . . 121 Treatment of Infectious Mimics of Viral Encephalitis . . . . 121 Treatment of Acute Viral Encephalitis . . . . 122 References . . . . 122

Contents

xi

7. Severe Head and Neck Infections in the Critical Care Unit . . . . . 125 Franco Paradisi and Giampaolo Corti Overview . . . . 125 Microbiology . . . . 126 Clinical Presentation of Deep Fascial Space Infections . . . . 132 Clinical Presentation of Other Cervical Infections . . . . 133 Diagnosis of Cervical Infections . . . . 135 Therapy of Cervical Infections . . . . 136 Clinical Presentation of Intracranial Suppurative Infections . . . . 140 Diagnosis of Intracranial Suppurative Infections . . . . 143 Therapy of Intracranial Suppurative Infections . . . . 146 References . . . . 151 8. Severe Community-Acquired Pneumonia in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Burke A. Cunha Introduction . . . . 157 Determinants of Severe CAP . . . . 158 Clinical Approach to Severe CAP . . . . 158 Clinical Approach to Severe Cap by Chest X-Ray Pattern and Degree of Hypoxemia . . . . 162 References . . . . 166 9. Nosocomial Pneumonia in the Critical Care Unit . . . . . . . . . . . . 169 Emilio Bouza, Almudena Burillo, and Marı´a V. Torres Overview . . . . 169 Pathogenesis . . . . 170 Microbiology . . . . 172 Risk Factors . . . . 174 Prevention . . . . 175 Clinical Presentation and Diagnostic Testing . . . . 181 Antimicrobial Treatment . . . . 186 References . . . . 191 10. Pleural Empyema/Lung Abscess in the Critical Care Unit John G. Bartlett Introduction . . . . 205 Empyema . . . . 205 Lung Abscess . . . . 210 References . . . . 215

. . . . . . 205

11. Infective Endocarditis and Its Mimics in the Critical Care Unit . . . 221 John L. Brusch Overview . . . . 221 Microbiology . . . . 224

xii

Contents

Microbial Pathogenicity . . . . 227 Clinical Presentation and Clinicopathological Correlations . . . . 230 PMIE . . . . 235 Differential Diagnosis Considerations . . . . 237 Diagnosis . . . . 242 Mimics of Endocarditis . . . . 243 Therapy . . . . 244 References . . . . 253 12. Acute Myocarditis and Acute Pericarditis in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 Jason M. Lazar, Diane H. Johnson, and Burke A. Cunha Acute Myocarditis . . . . 263 Acute Pericarditis . . . . 277 References . . . . 280 13. Central Intravenous Line Infections in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 Burke A. Cunha Introduction . . . . 283 Overview of CVC Infections . . . . 283 IV Line Infections . . . . 284 Septic Thrombophlebitis . . . . 285 S. aureus Acute Bacterial Endocarditis . . . . 286 References . . . . 287 14. Intra-Abdominal Surgical Infections and Their Mimics in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 Meghann L. Kaiser and Samuel E. Wilson Introduction . . . . 291 Tertiary Peritonitis . . . . 291 New Onset Peritonitis . . . . 294 Infectious Complications of Pancreatitis . . . . 300 Mimics of Abdominal Infection . . . . 301 De Novo Coincidental Intra-Abdominal Infection . . . . 302 References . . . . 302 15. Clostridium difficile –associated Diarrhea and Colitis in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Marci Drees and Sherwood L. Gorbach Overview . . . . 305 Microbiology . . . . 309 Clinical Presentation . . . . 310 Differential Diagnostic Considerations . . . . 311 Diagnosis . . . . 313 Treatment . . . . 314

Contents

xiii

Prevention . . . . 317 References . . . . 318 16. Severe Skin and Soft Tissue Infections in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321 Mamta Sharma and Louis D. Saravolatz Introduction . . . . 321 Microbial Flora . . . . 321 Classification of Skin and Soft Tissue Infections . . . . 322 Diabetic Foot Infection . . . . 337 Skin and Soft Tissue Infections in Injection Drug Users . . . . 339 Pyomyositis . . . . 339 Community-Acquired Methicillin-Resistant S. aureus . . . . 340 Summary . . . . 340 References . . . . 341 17. Infections in Patients on Steroids in the Critical Care Unit . . . . . . 347 John N. Sheagren Introduction . . . . 347 Historical Aspects of Steroids and Infections . . . . 347 Host Defenses/Immunity Against Infections . . . . 348 Microbial Defense by the Various Components of the Immune Systems . . . . 350 Major Effects of Steroids on Immunity . . . . 352 Types of Infections Expected in Patients on Steroids . . . . 354 Steroid Treatment of Infection . . . . 355 Conclusions . . . . 356 References . . . . 357 PART III: SPECIAL PROBLEMS IN THE CRITICAL CARE UNIT 18. Fever and Rash in the Critical Care Unit . . . . . . . . . . . . . . . . . . 361 Lee S. Engel, Charles V. Sanders, and Fred A. Lopez Introduction . . . . 361 Petechial and Purpuric Rashes . . . . 365 Maculopapular Rash . . . . 372 Diffuse Erythematous Rashes with Desquamation . . . . 377 Vesicular, Bullous, or Pustular Rashes . . . . 381 Nodular Rash . . . . 385 References . . . . 387 19. Tropical Infections in the Critical Care Unit . . . . . . . . . . . . . . . . 397 David R. Tribble and Kenneth F. Wagner Epidemiology of Infections in International Travelers . . . . 397 Malaria . . . . 399 Critical Care Infectious Disease Syndromes . . . . 402 References . . . . 410

xiv

Contents

20. HIV/AIDS in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . 417 Larry I. Lutwick Introduction . . . . 417 HAART Attacks . . . . 418 Heart Attacks . . . . 427 References . . . . 428 21. Infections in Cirrhosis in the Critical Care Unit . . . . . . . . . . . . . 433 Laurel C. Preheim Introduction . . . . 433 Role of the Liver in Host Defense Against Infection . . . . 434 Classification of Liver Disease Severity . . . . 434 Spontaneous Bacterial Peritonitis . . . . 434 Urinary Tract Infections . . . . 437 Bacteremia . . . . 437 Pneumonia . . . . 438 Other Infections . . . . 439 References . . . . 440 22. Infections Associated with Diabetes in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445 Larry I. Lutwick Infection and Diabetes . . . . 445 Life-Threatening Infection Characteristic of Diabetics . . . . 446 Life-Threatening Infection Characteristic of Diabetics in the Tropics . . . . 451 References . . . . 455 23. Infection in Organ Transplant Patients in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459 Patricia Mun˜oz, Almudena Burillo, and Emilio Bouza Introduction . . . . 459 Influence of the Type of Transplantation and of the Time after Transplantation . . . . 460 Most Common Clinical Syndromes . . . . 464 Management . . . . 480 Prevention . . . . 483 References . . . . 483 24. Infections in Asplenics in the Critical Care Unit . . . . . . . . . . . . . 497 Jihad Slim and Leon G. Smith Overview . . . . 497 Epidemiology . . . . 498 Microbiology . . . . 498 Clinical Presentation . . . . 500 References . . . . 503

Contents

xv

25. Infections in Burns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507 Steven E. Wolf, Basil A. Pruitt, and Seung H. Kim Introduction . . . . 507 Treatment of Burn Wound to Control Infection . . . . 508 Burn Wound Infection . . . . 512 Summary . . . . 523 References . . . . 523 26. Urosepsis in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . 527 Burke A. Cunha Introduction . . . . 527 Urosepsis . . . . 527 Antimicrobial Therapy . . . . 533 References . . . . 533 27. Infections Related to Bioterrorism . . . . . . . . . . . . . . . . . . . . . . . 535 David Schlossberg Overview . . . . 535 Microbiology . . . . 535 Clinical Presentation . . . . 536 Differential Diagnostic Considerations . . . . 541 Diagnosis . . . . 541 Therapy . . . . 545 Conclusion . . . . 550 References . . . . 551 PART IV: THERAPEUTIC CONSIDERATIONS 28. Antibiotic Dosing in Hepatic/Renal Insufficiency . . . . . . . . . . . . . 553 Damary C. Torres, Donna Sym, and April Correll Introduction . . . . 553 Antibiotic Dosing in Hepatic Failure . . . . 553 Antibiotic Dosing in Renal Insufficiency and Failure . . . . 555 Antibiotic Dosing in Intermittant Renal Replacement Therapy . . . . 561 Antibiotic Dosing in CRRT . . . . 566 Conclusion . . . . 571 References . . . . 571 29. Adverse Reactions to Antibiotics . . . . . . . . . . . . . . . . . . . . . . . . 575 Eric V. Granowitz and Richard B. Brown Introduction . . . . 575 Anaphylaxis . . . . 576 Nephrotoxicity . . . . 576 Hematological Adverse Reactions . . . . 579 Dermatological Toxicity . . . . 582

xvi

Contents

Neurotoxicity . . . . 583 Cardiotoxicity . . . . 585 Hepatotoxicity . . . . 586 Musculoskeletal Toxicity . . . . 586 Electrolyte Abnormalities . . . . 586 Fever . . . . 586 Antibiotic-Associated Diarrhea and Colitis . . . . 587 Antibiotic-Resistant Superinfections . . . . 589 Summary . . . . 589 References . . . . 589 30. Antibiotic Kinetics in the Febrile Multiple System Trauma Patient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595 Donald E. Fry Introduction . . . . 595 Normal Pharmacokinetics of Antibiotics . . . . 596 Pathophysiology of Injury and Fever . . . . 599 Clinical Data . . . . 600 Summary . . . . 606 References . . . . 606 31. Antibiotic Selection and Control of Resistance in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609 Burke A. Cunha Introduction . . . . 609 Antibiotic Selection in the CCU . . . . 609 Control of Resistance in the CCU . . . . 617 References . . . . 622 32. Antimicrobial Therapy in the Penicillin-Allergic Patient in the Critical Care Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625 Burke A. Cunha Introduction . . . . 625 Determining the Type of Penicillin Allergy . . . . 625 Penicillin Allergic Reactions . . . . 626 Cross Reactions Between Penicillins and b-Lactams . . . . 626 Carbapenems and Monobactams . . . . 627 Non–b-Lactam Antibiotics in Patients with Penicillin Anaphylactic Reactions . . . . 627 Conclusion . . . . 628 References . . . . 630 Index . . . . 633 About the Editor . . . . 663

Contributors

John G. Bartlett Department of Medicine, School of Medicine, Johns Hopkins University, Baltimore, Maryland, U.S.A. Emilio Bouza Clinical Microbiology and Infectious Diseases Department, Hospital General Universitario ‘‘Gregorio Maran˜o´n,’’ Universidad Complutense, Madrid, Spain Richard B. Brown Division of Infectious Disease, Baystate Medical Center and Tufts University School of Medicine, Springfield, Massachusetts, U.S.A. John L. Brusch Department of Medicine, Harvard Medical School, Cambridge, Massachusetts, U.S.A. Almudena Burillo Department of Clinical Microbiology, Hospital MadridMonteprı´ncipe, Madrid, Spain Brian W. Cooper Division of Infectious Disease, Hartford Hospital, Hartford, and University of Connecticut School of Medicine, Farmington, Connecticut, U.S.A. April Correll

Winthrop-University Hospital, Mineola, New York, U.S.A.

Giampaolo Corti Infectious Disease Unit, University of Florence School of Medicine, Florence, Italy Burke A. Cunha Infectious Disease Division, Winthrop-University Hospital, Mineola, and State University of New York School of Medicine, Stony Brook, New York, U.S.A. Marci Drees Department of Geographic Medicine and Infectious Diseases, Tufts-New England Medical Center, Boston, Massachusetts, U.S.A. Lee S. Engel Department of Medicine, Louisiana State University Health Science Center, New Orleans, Louisiana, U.S.A. Donald E. Fry Department of Surgery, University of New Mexico School of Medicine, Albuquerque, New Mexico, U.S.A. xvii

xviii

Contributors

Sherwood L. Gorbach Nutrition/Infection Unit, Department of Public Health and Family Medicine, Tufts University School of Medicine, Boston, Massachusetts, U.S.A. Eric V. Granowitz Division of Infectious Disease, Baystate Medical Center and Tufts University School of Medicine, Springfield, Massachusetts, U.S.A. Diane H. Johnson Infectious Disease Division, Winthrop-University Hospital, Mineola, and State University of New York School of Medicine, Stony Brook, New York, U.S.A. Meghann L. Kaiser Department of Surgery, University of California, Irvine School of Medicine, Orange, California, U.S.A. Seung H. Kim Burn Center, United States Army Institute of Surgical Research, San Antonio, Texas, U.S.A. Jason M. Lazar Department of Cardiology, State University of New York Downstate Medical Center, Brooklyn, New York, U.S.A. Fred A. Lopez Department of Medicine, Louisiana State University Health Science Center, New Orleans, Louisiana, U.S.A. Larry I. Lutwick Department of Infectious Diseases, VA New York Harbor Health Care System, and State University of New York Downstate Medical School, Brooklyn, New York, U.S.A. C. Glen Mayhall Division of Infectious Diseases and Department of Healthcare Epidemiology, University of Texas Medical Branch at Galveston, Galveston, Texas, U.S.A. Patricia Mun˜oz Clinical Microbiology and Infectious Diseases Department, Hospital General Universitario ‘‘Gregorio Maran˜o´n,’’ Universidad Complutense, Madrid, Spain Franco Paradisi Infectious Disease Unit, University of Florence School of Medicine, Florence, Italy Laurel C. Preheim Departments of Medicine, Medical Microbiology and Immunology, Creighton University School of Medicine, University of Nebraska College of Medicine, and VA Medical Center, Omaha, Nebraska, U.S.A. Basil A. Pruitt Division of Trauma and Emergency Surgery, Department of Surgery, University of Texas Health Science Center, and Burn Center, United States Army Institute of Surgical Research, San Antonio, Texas, U.S.A. Charles V. Sanders Department of Medicine, Louisiana State University Health Science Center, New Orleans, Louisiana, U.S.A.

Contributors

xix

Louis D. Saravolatz Division of Infectious Disease, Department of Medicine, St. John Hospital and Medical Center, and Wayne State University School of Medicine, Detroit, Michigan, U.S.A. David Schlossberg Infectious Disease Section, Department of Medicine, Temple University School of Medicine, Philadelphia, Pennsylvania, U.S.A. Mamta Sharma Division of Infectious Disease, Department of Medicine, St. John Hospital and Medical Center, and Wayne State University School of Medicine, Detroit, Michigan, U.S.A. John N. Sheagren Department of Internal Medicine, Advocate Illinois Masonic Medical Center, Chicago, Illinois, U.S.A. Jihad Slim Infectious Disease Division, Department of Medicine, Seton Hall P.G. School of Medicine, and St. Michael’s Medical Center, Newark, New Jersey, U.S.A. Leon G. Smith Infectious Disease Division, Department of Medicine, Seton Hall P.G. School of Medicine, and St. Michael’s Medical Center, Newark, New Jersey, U.S.A. Donna Sym College of Pharmacy and Allied Health Professions, St. John’s University, and North Shore University Hospital, Jamaica, New York, U.S.A. Damary C. Torres College of Pharmacy and Allied Health Professions, St. John’s University, Jamaica, and Winthrop-University Hospital, Mineola, New York, U.S.A. Marı´a V. Torres Clinical Microbiology and Infectious Diseases Department, Hospital General Universitario ‘‘Gregorio Maran˜o´n,’’ Universidad Complutense, Madrid, Spain David R. Tribble Department of Enteric Diseases, Infectious Diseases Directorate, Naval Medical Research Institute, Silver Spring, Maryland, U.S.A. Kenneth F. Wagner Independent Consultant, Infectious Diseases and Tropical Medicine, Islamorada, Florida, U.S.A. Samuel E. Wilson Department of Surgery, University of California, Irvine School of Medicine, Orange, California, U.S.A. Steven E. Wolf Division of Trauma and Emergency Surgery, Department of Surgery, University of Texas Health Science Center, and Burn Center, United States Army Institute of Surgical Research, San Antonio, Texas, U.S.A.

PART I: GENERAL CONCEPTS

1 Methicillin-Resistant Staphylococcus aureus/Vancomycin-Resistant Enterococci Colonization and Infection in the Critical Care Unit C. Glen Mayhall Division of Infectious Diseases and Department of Healthcare Epidemiology, University of Texas Medical Branch at Galveston, Galveston, Texas, U.S.A.

INTRODUCTION Methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE) are among the most common antibiotic-resistant nosocomial pathogens in health care in general and in critical care units (CCUs) in particular. Although discovered shortly after its introduction, resistance to methicillin was first reported in the United States in 1968 (1,2). Since then, MRSA have spread throughout the world and have continued to spread in the United States. In many healthcare facilities, 50% of S. aureus isolates are MRSA. In intensive care units (ICUs), MRSA now make up 60% of S. aureus isolates (3). As hospital-acquired methicillin-resistant S. aureus (HA-MRSA) continues to spread within healthcare facilities, sites where healthcare is delivered face a new threat from community-acquired methicillin-resistant S. aureus (CA-MRSA). These latter strains from the community first appeared in the 1990s and now have been detected throughout the United States and in many other countries throughout the world (4–12). Infections due to CA-MRSA occur in patients with no risk factors or recent contact with healthcare facilities. They commonly occur in healthy children and most commonly manifest as skin and soft-tissue infections (13–15). Most patients require treatment, and 23% to 29% have required hospitalization (14,15). The appearance of CA-MRSA raises concerns about an additional reservoir for MRSA for healthcare facilities. Indeed, reports have begun to appear of the introduction and spread of CA-MRSA in hospitals (16,17). Thus hospital epidemiologists and infection-control professionals will have to protect ICU patients from both HA-MRSA and CA-MRSA. VRE are resistant gram-positive cocci that have appeared more recently in hospitals and ICUs. VRE were first noted in November 1986 and reported in January 1988 (18). In July 1988, VRE colonization of hematology patients was reported from 1

2

Mayhall

Paris (19). In 1989, 0.3% of enterococci (0.1% in ICUs) isolated from patients in hospitals participating in the National Nosocomial Infection Surveillance (NNIS) system at the Centers for Disease Control and Prevention (CDC) were resistant to vancomycin (20). In 1993, 7.9% of enterococci isolated in NNIS system hospitals (13.6% in ICUs) were resistant to vancomycin. By 2003, 28.5% of enterococci isolated in NNIS system hospital ICUs were resistant to vancomycin (21). As normal flora, enterococci are not nearly as invasive as are S. aureus. Approximately 1 in 10 patients colonized with VRE develop infection (22), although this may vary with the degree of immunosuppression of the patients (23,24). The most serious infections with VRE are bacteremia, endocarditis, and meningitis. Urinary tract infections are less serious and easier to treat. Infections at other body sites are difficult to document, because VRE isolated from other sites frequently represent colonization and not infection (25,26).

METHICILLIN-RESISTANT S. AUREUS Types of MRSA Nosocomial Methicillin-Resistant S. aureus Nosocomial methicillin-resistant S. aureus (NA-MRSA) first appeared in the United States in 1968 (2). It has spread across the United States over the last three-and-ahalf decades by lateral transfer among hospital patients and by transfer of patients between hospitals and between hospitals and long-term care facilities. Most circulating strains of NA-MRSA appear to have originated from two or three clones of MRSA (27,28). Methicillin resistance and resistance to all betalactam antibiotics are conferred by the staphylococcal cassette chromosome mec (SCCmec), which carries the mecA gene that encodes a protein designated ‘‘penicillin-binding protein 2a’’ or ‘‘penicillinbinding protein 20 .’’ These altered penicillin-binding proteins bind betalactam antibiotics poorly, permitting cell wall synthesis to continue in the presence of these antimicrobial agents. There are three types of SCCmec in HA-MRSA, types I, II, and III (4,29). Type I contains no additional resistance determinants, but types II and III contain resistance determinants in addition to mecA; these additional genetic elements account for the antimicrobial resistance to many antibiotics in addition to the betalactam agents. The three SCCmec types contained in NA-MRSA have an identical chromosomal integration site and the cassette chromosome recombinase genes, which are responsible for horizontal transfer of SCCmec (4). Thus, NA-MRSA are resistant to many antibiotics and have a selective advantage as they are spread among patients by the hands of personnel and other contaminated surfaces. The presence of underlying diseases and multiple types of instrumentation and procedures predisposes patients to colonization and infection by the multiply resistant strains of NA-MRSA. Community-Acquired MRSA. CA-MRSA have appeared gradually over about the last 15 years. Early on there was uncertainty about the origin of CA-MRSA, and it was unclear whether CA-MRSA were different from NA-MRSA. Some investigators believed that most of the CA-MRSA infections could be traced back to some previous contact with the healthcare system. More recently, it has become clear that these infections occur in young healthy persons with no recent healthcare contacts

MRSA/VRE Colonization and Infection in the Critical Care Unit

3

and no risk factors for NA-MRSA. It has also become clear that CA-MRSA have evolved in the community through an evolutionary pathway entirely separate from NA-MRSA. It appears that all four of the SCCmec types have risen from Staphylococcus sciuri, the most ubiquitous and ancient species of Staphylococcus (30). Due to their large size, SCCmec types I, II, and III have rarely been transferred to the cells of methicillin-susceptible S. aureus (MSSA). On the contrary, CA-MRSA has an SCCmec type IV that is small enough to be transferred between cells by transduction or phage-mediated transformation (30,31). There is some evidence that transfer of type IV SCCmec from CA-MRSA to MSSA can occur (29). Given that many infections caused by CA-MRSA are treated in hospitals and other healthcare facilities, there must be some concern that CA-MRSA may become another type of MRSA in hospitals. In addition to infections, it is likely that patients admitted to hospitals for a variety of indications will be colonized with CA-MRSA. In addition to adding to the burden of MRSA in the hospital, CA-MRSA appear to be more virulent than NA-MRSA. The MW2 strain of CA-MRSA, a common strain in the United States, has 18 toxins which were not found in five comparative S. aureus genomes (32). The majority of CA-MRSA contain the genetic element for the Panton-Valentine leukocidin. This toxin has been associated with necrotizing pneumonia in healthy children (6). The MW2 strain of CA-MRSA contains genes for 11 exotoxins and four enterotoxins. All of these toxins are superantigens (32). CA-MRSA may also contain genes for exfoliative toxins and for hemolysins (33). CA-MRSA most commonly cause skin and soft tissue infections in persons with no risk factors for NA-MRSA. However, they may cause severe disease, and hospital patients may be at particularly high risk for serious disease. It is very important that infection control programs be on guard for ingress of CA-MRSA into hospitals, and this is particularly true for ICUs.

Types of Infections Caused by MRSA Infections Caused by NA-MRSA Adult ICUs. Bacteremia and pneumonia are the most common NA-MRSA infections encountered when all types of ICUs are considered (34–39). Other NA-MRSA infections reported include urinary tract infections (34,35), empyema (35), and bacteremia associated with hemofiltration (38). Surgical site infections due to NA-MRSA are reported from ICUs that care for surgical patients, although most all of these infections were acquired in the operating room and not in the ICU (35,36). Neonatal ICUs. NA-MRSA are recovered from many more sites of infection in patients in neonatal intensive care units (NICUs) compared with patients in adult ICUs. As is the case in adult ICUs, reports on sites of infection due to NA-MRSA in neonates are from publications of outbreak investigations. Table 1 shows the sites of infection due to NA-MRSA reported from outbreaks in NICUs. Infections Caused by CA-MRSA Adult ICUs. To date, all cases of CA-MRSA acquired in the hospital by adults have been reported from Australia (45–47). There were no reports of

4

Mayhall

Table 1 Sites of Infection Due to Nosocomial Methicillin-Resistant Staphylococcus aureus in Patients in Neonatal Intensive Care Units Sites of infection Bacteremia, primary Pneumonia Skin and soft tissue abscess Peritonitis or necrotizing enterocolitis Ventriculitis or meningitis Osteomyelitis or septic arthritis Urinary tract infection Eye infection Wound infection Endocarditis Thrombophlebitis Ear, nose and throat infection Omphalitis Source: From Refs. 40–44.

outbreaks in the ICUs of these hospitals. None of the isolates were tested for type of SCCmec or for Panton-Valentine leukocidin or other toxins often found in CA-MRSA. The strains of CA-MRSA were isolated from 24% to 42% of inhabitants in two communities remote from the urban area where patients from these communities were hospitalized. Only one of these reports provided limited information on the sites of infections (46). In the latter report, there were 19 episodes of bacteremia in 16 patients. Neonatal ICUs. Two outbreaks due to CA-MRSA have been reported from NICUs (17,48). In one outbreak, the isolates were identified as CA-MRSA by detection of a type IV SCCmec (17). However, virulence factors frequently found in CA-MRSA, including the element coding for Panton-Valentine leukocidin, were not detected in the outbreak strain. The second outbreak in an NICU was stated to be due to CA-MRSA, but no testing for the type of SCCmec or virulence factors was done (48). The mother of the index case had had contact with the healthcare system, and the antibiogram of the isolates suggested that they were likely NA-MRSA. An outbreak has also been reported in a newborn nursery and associated maternity units (49). The isolates from this outbreak were shown to have the type IV SCCmec and genes for Panton-Valentine leukocidin and staphylococcal enterotoxin K. Epidemiology of NA-MRSA Infections in Critical Care Epidemiology of NA-MRSA Adult ICUs. The risk for adult patients who are culture-negative for NA-MRSA on admission to an ICU, where NA-MRSA is endemic, for acquiring NA-MRSA ranges between 4.5% and 11.7% for cumulative incidence (36,50) and between 7.9 and 9.9 per 1000 patient days for incidence density (51,52). In one study, it was observed that NA-MRSA was acquired at about 1% per day in the first week after admission and then at 3% per day thereafter (38).

MRSA/VRE Colonization and Infection in the Critical Care Unit

5

Sources of NA-MRSA. The sources of NA-MRSA include colonized or infected patients, colonized or infected healthcare workers (HCWs), and contaminated environmental surfaces. One of the best indications of the importance of colonized and infected patients as an important source of NA-MRSA is the significant relationship between colonization pressure and acquisition of NA-MRSA colonization or infection by patients who have no colonization or infection due to NA-MRSA at the time of admission to an ICU (50). Colonization pressure is defined as the number of patient days for patients with cultures positive for NA-MRSA divided by the number of total patient days (53). Colonization pressure can be calculated for any day or for a given period of time. The most common site of MRSA colonization is the external nares (35,54,55). The second most common site of colonization is skin and soft tissue other than surgical sites (34%) (54). Other sites of colonization include rectal (11–28.9%), respiratory tract (11%), and urinary tract (6%) (35,54,55). Another source of NA-MRSA is colonized or infected healthcare personnel. Acquisition of NA-MRSA in an ICU from a respiratory therapist with chronic sinusitis due to NA-MRSA has been reported, as well as surgical site infections due to colonization of the external nares and an area of dermatitis on the hand of a surgeon (56,57). The surgical site infections caused by the colonized surgeon were initiated at the time of surgery but became manifest postoperatively in the ICU. HCWs often become colonized with NA-MRSA from patient contacts when providing healthcare but are not often implicated in transmission to patients. To implicate a colonized HCW as a source for colonization or infection of patients, it is first necessary to epidemiologically establish an association between contact with the colonized or infected HCW and acquisition of NA-MRSA by patients. Then it is necessary to prove that the strain from the HCW and the patient is the same using molecular techniques such as pulsed-field gel electrophoresis (PGFE) after restriction endonuclease digestion of genomic DNA. Contaminated surfaces of equipment and environmental surfaces appear to make up another source of NA-MRSA for transmission to patients (58,59). NA-MRSA has been recovered from cultures of computer terminals, the floor next to the patient’s bed, bed linens, patient gowns, over-bed tables, blood pressure cuffs, bedside rails, infusion pump buttons, door handles, bedside commodes, stethoscopes, and window sills. In the latter study, 27% of 350 environmental surface cultures yielded NA-MRSA (59). It has also been shown in in vitro studies that outbreak isolates of NA-MRSA survive at significantly higher concentrations and for longer periods of time on an inanimate surface than do sporadic NA-MRSA isolates (60). Thus, environmental contamination is likely another important source for transmission of NA-MRSA to patients. Mode of Transmission of NA-MRSA. The most common mode of transmission of NA-MRSA to patients is by indirect contact. Several studies have shown that NA-MRSA is frequently transmitted to the hands and clothing of HCWs from colonized or infected patients. Two studies have shown that NA-MRSA can be recovered from 14% to 17% of HCWs’ hands after patient contact (61,62). Another study showed that 7 out of 12 (58%) nurses who cared for patients with NA-MRSA in a wound or urine had NA-MRSA on their gloves, recoverable by direct plating to solid media (59). Culture of 13 of 20 (65%) nurses’ uniforms or gowns who cared for these same patients yielded NA-MRSA. When cultures were taken from gloves of 12 personnel who touched only environmental surfaces in the rooms of these patients, five (42%) had NA-MRSA recovered on culture. Arbitrary-primed

6

Mayhall

polymerase chain reaction (PCR) typing demonstrated that isolates recovered from patients and environment had very similar banding patterns (59). Although additional studies are needed, data continue to accumulate in support of indirect transfer of NA-MRSA to patients from contaminated hands and clothing of HCWs. NA-MRSA also appear to have an advantage over MSSA in colonizing patients after transmission (63). During an epidemic of NA-MRSA colonizations and infections in a surgical ICU, 23 patients were exposed to six patients admitted to the ICU with NA-MRSA colonization. PFGE of isolates showed that all secondary cases had NA-MRSA PFGE patterns identical to the PFGE patterns of the strain recovered from the patients to whom they were exposed. None of the PFGE patterns of the isolates of MSSA cultured from patients and HCWs were the same. The authors concluded that NA-MRSA may have spread more easily between patients due to selection through antibiotic pressure. Airborne transmission of NA-MRSA may occur, but the importance of this route of transmission has not been established. The CDC has not recommended airborne precautions for patients with NA-MRSA colonization or infection (64). Theoretically, NA-MRSA could be transferred by the airborne route after aerosolization from contaminated environmental surfaces or by aerosolization from nasal carriers. One study has shown that NA-MRSA can be aerosolized from environmental surfaces, i.e., changing bed sheets (65). Molecular typing showed that environmental isolates and patient isolates were identical. However, the authors did not investigate other possible routes of transmission of NA-MRSA to the patients. Several studies have been published on the dissemination of S. aureus from the upper respiratory tracts of HCWs. To the author’s knowledge, no such studies have been published on dissemination of NA-MRSA from HCWs. One study has epidemiologically implicated a HCW with chronic sinusitis and nasal colonization with S. aureus in spread of S. aureus to patients. The relationship was confirmed by molecular typing (56). There appears to be a strong relationship between shedding of S. aureus by HCWs and having a viral upper respiratory tract infection (66,67). In one study, nasal carriers of S. aureus who volunteered were experimentally infected with rhinovirus (67). Investigators were able to quantify the S. aureus colony-forming units (CFU) released into the air under varying conditions including type of clothes worn and whether or not a mask was worn. They documented that the S. aureus released into the air was from the experimentally infected volunteers by molecular typing. Studies on airborne dissemination of NA-MRSA using these techniques are needed. Risk Factors for Acquisition of NA-MRSA. Risk factors for acquisition of NAMRSA in ICUs vary depending on the type of ICU. Risk for NA-MRSA colonization/ infection identified in recent well-designed studies making use of multivariable analysis is shown in Table 2. Neonatal ICUs. The epidemiology of NA-MRSA colonization and infection has been less well studied in NICUs than in adult ICUs. Few, if any, reports on outbreaks of NA-MRSA in NICUs published in the 1990s and up to the present have included data on the risk of acquisition of NA-MRSA during outbreaks or analytic epidemiologic studies to identify risk factors for acquisition. One study provided time-and-intensity-of-care-adjusted incidence density for infections. In the intensive care section of the unit this incidence density was 0.73 infections/1000 patient-care hours (40). In the intermediate-care area the incidence density was 0.62 infections/1000 patient-care hours. There are no data on the rate of acquisition of NA-MRSA colonization.

Medical Interdisciplinary

Trauma

Merrer et al. (50) Grundmann et al. (52)

Marshall et al. (68)

Previous admission to the ICU Previous admission to trauma/ orthopedics ward Previous admission to the neurology/endocrinology/ rheumatology/renal ward LOS more than three days prior to admission to the ICU Being a trauma patient LOS 2–7 days in the ICU LOS more than 7 days in the ICU Weekly colonization pressure > 40% Clustered cases Days of staff deficit Sporadic cases Urgent/emergency admission APACHE II score at 24 hr Bronchoscopy Laparotomy Motor vehicle accident Ticarcillin–clavulanic acid Glycopeptide

Risk factors

0.011 0.044 0.009

(1.328–9.209) (1.002–1.147) (1.38–9.84) (1.4–28.9) (1.2–93.7) (1.3–15.0) (1.7–21.0)

3.50 1.07 3.68 6.3 10.4 4.5 5.9

< 0.0001 0.001

(1.8–8.7) (1.4–86) (14.5–833) (1.7–20.1)

p Value

1.05 (1.020–1.084)

3.9 11.1 109.8 5.8

8.6 (4.4–16.9)

2.6 (1.0–6.9)

3.3 (1.7–6.6) 2.9 (1.2–7.2)

Adjusted odds ratio (95% CI)

Abbreviations: ICU, intensive care unit; LOS, length of stay; APACHE, acute physiology and chronic health evaluation.

Medical surgical

Type of ICU

Marshall et al. (36)

Publications

Table 2 Risk Factors for Acquisition of Nosocomial Methicillin-Resistant Staphylococcus aureus in Adults

MRSA/VRE Colonization and Infection in the Critical Care Unit 7

8

Mayhall

There are few data on the source of NA-MRSA in NICUs. In one recent study, patients would have to be presumed to be the source of NA-MRSA, as personnel or the environment could not be implicated (42). In another study based on molecular typing, environmental cultures were all negative and a HCW was thought to have transferred the NA-MRSA outbreak strain from an adult hospital (44). However, the HCW was not epidemiologically implicated as the source. In all of the latter studies, transmission between patients by the hands of HCWs is suggested (40,42,44). No case–control studies to identify risk factors for colonization or infection with NA-MRSA in NICUs have been published to the author’s knowledge. Using a different approach, one study implicated overcrowding and understaffing as risk factors for acquisition of NA-MRSA colonization or infection (40). Epidemiology of CA-MRSA Adult ICUs. Although outbreaks of CA-MRSA infections have been described in hospitals in Australia, there were no reports of such outbreaks in ICUs (45–47). There have been no reports of infections due to CA-MRSA in adult ICUs in the United States. Neonatal ICUs. There are three published reports of transmission of CAMRSA in NICUs (17,48,69). In one report, the strain was identified as CA-MRSA by recovery from both the mother and neonate within the first 48 hours after admission (48). However, the isolate was susceptible only to gentamicin, rifampin, and vancomycin, and the mother had had contact with the healthcare system for prenatal care. The isolate was not tested for either the SCCmec type or the gene that encodes for Panton-Valentine leukocidin. Strains from the other two studies were both tested for the SCCmec type and Panton-Valentine leukocidin (17,69). Isolates from these two investigations had SCCmec type IV identified, but one did not carry the gene for Panton-Vanentine leukocidin (17) and the other was not tested for the latter virulence factor (69). One study provided an incidence rate of 18.5 cases per 1000 hospitalized neonates (17). There are no published data on modes of transmission or risk factors for acquisition of CA-MRSA in NICUs. Although there are few published epidemiologic data on the spread of CA-MRSA in NICUs, it is clear that CA-MRSA may enter NICUs and cause outbreaks with resultant colonization and infection of neonates. It is likely that CA-MRSA will continue to enter many areas of hospitals, and more definitive studies will be needed to better understand how to prevent entry of CA-MRSA and to control it once present in healthcare facilities. Prevention and Control of MRSA in ICUs Prevention of MRSA transmission and control of ongoing dissemination among patients receiving healthcare require a number of preventive and control measures. The approach to control is similar for adult and neonatal patients and for NA-MRSA and CA-MRSA. Differences for adults versus neonates and for NA-MRSA versus CA-MRSA will be noted. Screening Patients on Admission and During Hospitalization The most important measures for control of MRSA in ICUs are active surveillance for patients infected or colonized with MRSA at the time of admission followed by prompt isolation of those patients identified as colonized or infected and weekly

MRSA/VRE Colonization and Infection in the Critical Care Unit

9

cultures for patients remaining in the ICU to detect acquisition of MRSA from patients who may have escaped detection on admission, from colonized or infected HCWs, or from contaminated environmental surfaces (34,44,63,70–84). It is important to identify every colonized patient so that all colonized as well as all infected patients can be placed on contact precautions. Surveillance cultures for MRSA should always include samples from the anterior nares (70). Screening patients for colonization with MRSA has been done by taking swab samples from the anterior nares and other sites of possible MRSA colonization, such as the oropharynx, axilla, inguinal area, perirectal areas, and from open wounds and skin eruptions. Samples were then inoculated to broth or solid media containing antibiotics or other agents to select out MRSA. Although effective, results are not immediately available due to the delay for incubation and identification of isolates. More rapid techniques for detection of MRSA based on the PCR have been developed and published (85). Such techniques permit detection of MRSA from swab specimens within two hours. Screening for MRSA colonization and infection on admission is particularly important for patients admitted from other hospitals, from long-term care facilities, or who have been hospitalized in the past year. Although it is not yet clear as to the impact of CA-MRSA on the influx of MRSA into hospitals, this potential reservoir for MRSA must be kept in mind. It may be necessary to screen everyone entering the hospital from the community regardless of whether they have one of the above-mentioned risk factors for MRSA colonization or infection. Barrier Precautions Gloves should be worn before entry of HCWs into rooms of patients isolated for MRSA (70). There is good evidence that HCWs acquire MRSA on gloved and ungloved hands when in contact with patients colonized or infected with MRSA (61,62). Hands should be washed before and after glove use. Gowns should be worn on entry into the room except when there will be no contact between the HCW and the patient or between the HCW and environmental surfaces (70). Studies have shown that the clothing of HCWs becomes contaminated after contact with patients and patient-care surfaces (59,86). Whether or not masks are needed for contact precautions for MRSA is controversial. The CDC has not recommended that masks be used for isolation of patients colonized or infected by MRSA (64). Masks are recommended by the Society for Healthcare Epidemiology of America (SHEA) Guidelines for preventing nosocomial transmission of multidrug-resistant strains of S. aureus and Enterococcus (70). However, the recommendation is categorized as a type II. Definitive studies are needed to determine whether or not masks are needed for isolation of patients with MRSA colonization or infection. Decontamination of the Environment There is growing evidence that the environment may be an important source for MRSA for patient colonization and infection (59,87,88). One study has shown that strains of MRSA survive for about 7 to 10 months on glass surfaces (60). It was also shown that outbreak strains of MRSA survived longer than sporadic strains. There is evidence that enhanced disinfection is an important measure for controlling epidemic MRSA (89,90). Thus, attention should be paid to thorough cleaning and disinfection of environmental surfaces in patient rooms and other areas where patients receive care.

10

Mayhall

Hand Hygiene Hand hygiene is very important in conjunction with barrier precautions in preventing the spread of MRSA between patients and from patients to HCWs (70). Hand hygiene practices have been suboptimal for many years, and efforts to improve them have had little impact on compliance rates, which average about 40%. Risk factors for poor compliance include being a physician or a nursing assistant, working in an ICU, working during weekdays performing activities with a high risk for transmission, and having many opportunities for hand hygiene per hour of patient care (70). Most of these risk factors for poor hand hygiene are commonly present in ICUs. HCWs must be taught to decontaminate their hands with an antisepticcontaining agent (an alcohol-based hand rub or a hand washing preparation containing an antiseptic agent). If hands are visibly soiled with urine, feces, blood, or other body fluids, they must be washed with soap and water followed by application of an alcoholbased hand rub or washed with soap containing an antiseptic. Hands must be decontaminated before and after contact with each patient. This includes decontamination by washing with an antimicrobial soap or application of an alcohol-based hand rub after removal of gloves (91). HCWs should be strongly encouraged to apply moisturizing hand lotions, but it is important to establish that such preparations are compatible with the cleansing products and glove materials used by the HCWs. HCWs must be thoroughly educated about microbial contamination of their hands and why hand hygiene is important. Hand hygiene should be monitored and feedback should be given to HCWs about their performance on a continuous basis. It is unlikely that occasional feedback will change hand-hygiene practice. Decolonization of Patients Who Are Carriers of MRSA Decolonization of patients as a way to prevent and control outbreaks of colonization and infections due to both MRSA and MSSA has been studied for decades. In spite of the introduction of mupirocin as one of the most potent topical antistaphylococcal antibiotics discovered to date, decolonization of patients colonized with MRSA remains a challenge (92). In a number of studies, patients often become recolonized with the same or a different strain of MRSA. Few randomized controlled clinical trials with long-term follow-up (12 weeks after intranasal application of mupirocin) have been conducted. Decolonization is often attempted using a combination of mupirocin applied to the nares and showers with an antiseptic agent such as chlorhexidine. Very little published data suggest that chlorhexidine baths may add to the efficacy of mupirocin (93). One of the major problems in the use of mupirocin for decolonization of patients, in addition to failure to maintain long-term decolonization, is development of resistance (94). Resistance is particularly likely to develop with extensive use such as application to wounds. Resistance to mupirocin after use for treatment of both colonization and infection can be effectively controlled by limiting its use to the treatment of colonization (94). Use of mupirocin for decolonization of patients in ICUs must be very judicious. Several of the risk factors for failure are present in many ICU patients (92). These include (i) colonization of multiple body sites; (ii) chronic nonhealing wounds; and (iii) the presence of colonized foreign bodies such as tracheostomy tubes or gastrostomy tubes. Treatment for colonization should be limited to the nares. Attempts at decolonization of patients with colonization at multiple body sites, with chronic nonhealing wounds, and the presence of foreign bodies should be avoided. If mupirocin is

MRSA/VRE Colonization and Infection in the Critical Care Unit

11

used on multiple patients over long periods of time (months), MRSA isolates from patients should be tested for susceptibility to mupirocin. Another approach to decolonization of MRSA carriers has been instillation of vancomycin into the gastrointestinal tract by way of a nasogastric tube. In a recent study, the ICU patients had surveillance cultures of throat and rectum for MRSA over an eight-month period (95). The patients were part of a study of prevention of infection in mechanically ventilated patients. The patients were receiving oral antimicrobial agents for selective decontamination of the digestive tract. The authors designed a study to determine whether oral administration of vancomycin could eliminate MRSA from the intestinal tract. The study was not randomized and did not have concurrent controls. The authors noted a significant decrease in MRSA infections in the treated group compared with the historical group. They were able to show elimination of MRSA from the gastrointestinal tract based on rectal swab cultures. The weaknesses of the study included nonrandomization, the use of historic controls, and the simultaneous administration of other oral antimicrobial agents. The strengths included eradication of gastrointestinal carriage of MRSA and the careful monitoring of vancomycin resistance in MRSA and enterococci. No resistance was detected in many isolates of MRSA and enterococci tested for vancomycin susceptibility during the study. The authors also noted that by eradicating rectal carriage with vancomycin and preventing infection, they administered only 25% as much vancomycin to the group given oral vancomycin prophylaxis as was needed to treat the infections in the control group. Additional studies are required to better define the role of oral vancomycin for decolonization of the gastrointestinal tract, but this modality of decolonization appears to be of potential benefit and is worthy of further investigation. Decolonization of patients in NICUs is similar to that in adult ICUs but has not been as well studied. In one report of a MRSA outbreak, four patients were treated with nasal mupirocin three times a day for five days and bathed with diluted (1:10); 4% chlorhexidine gluconate once daily for three days (17). Two of the four neonates were successfully decolonized and two remained colonized with MRSA. The latter two were decolonized after the regimen was repeated. In a report of a second outbreak, colonized neonates were treated with mupirocin twice daily to the anterior nares and the umbilical area for seven days (96). The authors did not report the results of their decolonization regimen. In an account of a MRSA outbreak in an NICU, one control measure was application of triple dye to the umbilical area of the patients (40). This was one of several control measures implemented. Other control measures instituted included reducing overcrowding and understaffing and placing an infection control nurse in the NICU. Because all of these control measures were implemented at the same time, it was not possible to determine what effect the triple dye had in controlling the outbreak. Decolonization of Healthcare Workers Who Are MRSA Carriers Decolonization of HCWs is necessary when they have been epidemiologically implicated in the transmission of MRSA to patients from a colonized body site, which is most often the nose. Eradication of MRSA carriage from HCWs has been shown to help control outbreaks (56). For MRSA, mupirocin will decolonize the external nares effectively 91% of the time, although recolonization may occur in about one quarter of individuals so treated within four weeks (97). It has also been shown that

12

Mayhall

decolonization of HCWs with nasal carriage of MSSA results in a substantial decrease in hand carriage (98). Temporary decolonization of most of the colonized HCWs in an ICU for a few weeks may help control an outbreak. Although there are few data on decolonization of HCWs carrying MRSA, it is likely that mupirocin will eradicate MRSA from the nares and hands of HCWs. A second area where HCWs may be colonized with MRSA is at the site of dermatitis on their hands or forearms. It is important that hands and forearms of HCWs be examined and areas of dermatitis be cultured during an outbreak investigation. Other sites of colonization or infection are less common but may have to be sought if epidemiologically indicated. Table 3 lists the control measures for MRSA in ICUs. Cost Effectiveness of MRSA Control One study of the cost-effectiveness of MRSA control in an medical intensive care unit (MICU) has concluded that identification of patients who are carriers of MRSA on admission and during hospitalization and isolating of these carriers is cost effective (34). In spite of an ongoing MRSA carriage prevalence in admitted patients of 4%, the authors were able to reduce the incidence of ICU-acquired MRSA infection and colonization by fourfold. They observed that costs for single-room isolation of patients were $1480 and that the extra cost of an MRSA infection was $9275. They estimated that control was cost effective when MRSA carriage on admission is between 1% and 7% and when the MRSA transmission rate from colonized to isolated patients is at least fivefold less than to patients not isolated. Additional studies are needed on the cost effectiveness of MRSA control. VANCOMYCIN-RESISTANT ENTEROCOCCI Mechanism of Resistance Although there are many species of Enterococcus, relatively few species make up the VRE that cause endemic and epidemic nosocomial colonization and infection in healthcare facilities. The most important species are E. faecium and E. faecalis. Two other species, E. gallinarum and E. casseliflavus, are motile and display intrinsic vancomycin resistance (99). Vancomycin resistance in enterococci is mediated by the production of D-Alanine:D-Alanine ligases of altered substrate specificity (100). The most common ligases with altered substrate specificity are vanA and vanB. Both of these ligases condense D-Ala with D-Lac (lactate). Vancomycin does not bind to D-Lac, thus permitting cell wall synthesis to continue. The vanA trait is carried on a transposon, Tn1546. This transposon is most often carried on a plasmid and can be transferred to other gram-positive cocci. The genes that code for both vanA and vanB are similar. The vanB genes are carried on a large mobile element found on the chromosome. The vanB trait can be transferred to other enterococci (99). VRE containing the vanA ligase are resistant to vancomycin and another glycopeptide, teicoplanin, whereas vanB isolates are resistant to vancomycin but are susceptible to teicoplanin. Enterococci carrying vanA have minimal inhibitory concentrations (MICs) to vancomycin of >64 mg/mL, whereas isolates with vanB have MICs to vancomycin of 16 to >1000 mg/mL (99). Other types of ligases with altered substrate specificities are vanC [D-Ala-D-Ser (serine)], vanD (D-Ala-D-Lac), and vanE (D-Ala-D-Ser). The vanE genes are found on

MRSA/VRE Colonization and Infection in the Critical Care Unit

13

Table 3 Control Measures for Methicillin-resistant Staphylococcus aureus in Intensive Care Units Measure Culture all patients on admission and weekly while in the ICU until they become positive for MRSA or they are discharged

Place patients with MRSA infection and colonization on contact precautions

Practice hand hygiene after leaving room

Culture environmental surfaces to assess extent of contamination with MRSA

Decontaminate environmental surfaces often enough to keep them free of MRSA

Determine what sites to clean and the frequency of cleaning based on environmental culture data Attempts at decolonization of patients with MRSA should be done only under the supervision of infection control staff

Comments Use selective culture media Always take cultures from the external nares Culture wounds and skin eruptions Consider perirectal cultures if other sites are negative Flag patients’ charts or flag patients in the hospital computer system who are MRSA positive Place patients flagged for MRSA on contact precautions on admission Wear gloves to enter the room Wear a gown for contact with the patient or environment Use of a mask is optional Remove gloves and gown prior to leaving the room Wash hands with soap containing an antiseptic or apply an alcohol hand rub If hands are visibly soiled, wash with a soap containing an antiseptic or wash with plain soap followed by application of an alcohol hand rub Obtain specimens with sterile swabs moistened with sterile saline without bacteriostatic agents Use selective culture media to maximize efficiency of laboratory identification of MRSA Thoroughly clean surfaces followed by application of a hospital-grade disinfectant Culture environmental surfaces to determine effectiveness of cleaning and disinfection methods Do not use phenolic disinfectants in NICUs for environmental decontamination

Mupirocin is the agent of choice Follow the manufacturer’s instructions for use Decolonization should be attempted for nasal colonization only Attempts at nasal decolonization should not be done for patients with the following conditions: Colonization of multiple body sites Chronic nonhealing wounds (Continued )

14

Mayhall

Table 3 Control Measures for Methicillin-resistant Staphylococcus aureus in Intensive Care Units (Continued ) Measure

Comments

Presence of colonized foreign bodies such as tracheostomy tubes or gastrostomy tubes Take cultures after treatment for decolonization and 12 wks later Nasal decolonization is the same in NICUs Mupirocin should be applied to the external Healthcare workers who have nasal nares according to manufacturer’s colonization with MRSA and who have instructions been epidemiologically implicated in Follow up cultures of the external nares transmission to patients should be should be taken after therapy and again at furloughed from patient care and treated 2, 6, and 12 wks to detect relapse or with mupirocin for decolonization recolonization When decolonization is unsuccessful on the first attempt, retreatment may be successful Sites of infection or colonization should be When healthcare workers are infected with culture negative before the healthcare MRSA or have colonization of dermatitis, worker returns to patient care they should be furloughed from patient care and treated for infection or dermatitis until the condition clears Abbreviations: MRSA, methicillin-resistant Staphylococcus aureus; ICU, intensive care unit; NICUs, neonatal intensive care units.

the chromosomes of E. gallinarum and E. casseliflavus. These latter species have intrinsic low-level resistance to vancomycin (8–16 mg/mL). More recently, it has been discovered that E. faecium strains of VRE have acquired genes that appear to code for two virulence factors (101,102). The esp gene was found only in outbreak strains of E. faecium on three continents and not in nonepidemic isolates and isolates from healthy individuals or farm animals (101). Isolates carrying the esp gene seem to be associated with in-hospital spread and possibly with increased virulence. The hylEfm gene is found primarily in vancomycinresistant E. faecium in nonstool cultures obtained from patients hospitalized in the United States (102). This observation suggests that specific E. faecium strains may contain determinants that are associated with clinical infections. The appearance of virulence determinants in microorganisms that were considered nonvirulent normal flora in the past makes control of VRE even more urgent than when the only concern was resistance to glycopeptides. Types of Infections Caused by VRE Adult ICUs The most important type of infection caused by VRE is bacteremia. Such infections are usually related to intravascular catheters (103–109). Mortality due to VRE bacteremia has not been studied extensively. One study concluded that VRE bacteremia had a negative impact on survival (107). The best study was a historical cohort study that found an attributable mortality of 37% (95% CI 10–64%) (106). Nosocomial meningitis has been reported rarely (110,111). VRE is frequently cultured

MRSA/VRE Colonization and Infection in the Critical Care Unit

15

from urine, but only about 13% of patients with positive urine cultures have a urinary tract infection. Bacteremia from the infected urinary tract occurs but is uncommon (112). A univariate analysis of patients with and without a urinary tract infection revealed a significant relationship between having a malignancy and a urinary tract infection (112). Neonatal ICUs As in adults, neonates may also develop serious infections caused by VRE (113–115). The most common infection is bacteremia. Meningitis due to VRE has been reported in neonates, and two cases of VRE meningitis developed in patients after ventriculoperitoneal shunt placement (114). Urinary tract infection and lower respiratory tract infection with VRE has also been reported (114). Similar to adult patients, only about 1 in 10 colonized patients develop infection.

Epidemiology of VRE in ICUs Sources of VRE The main source/reservoir for VRE in hospitalized patients is the gastrointestinal tract (116–119). The first sites from which VRE are recovered on culture in newly colonized patients 86% of the time are the rectum or groin (116). Rectal cultures for VRE remain positive 100% of the time while patients are hospitalized. Gastrointestinal colonization may be very prevalent in ICU patients even in the absence of an outbreak (118). Patients with gastrointestinal colonization with VRE have very high concentrations of VRE in stool (median 108 CFU/g) (117). VRE are the predominant aerobic microorganisms in the gastrointestinal tracts of colonized patients, outnumbering gram-negative bacilli and vancomycin-susceptible enterococci. Given the high concentrations of VRE in stool, it is not surprising that many body sites in the patient carrying VRE become colonized (116). Transmission of VRE in the ICU Transmission of VRE to patients is by indirect contact with the hands of HCWs and fomites. There is no evidence that VRE are spread by the airborne route. Four studies show that gloved hands in contact with colonized patients and their environments become culture positive for VRE (120–123). When patients have diarrhea, the likelihood of HCWs picking up VRE on their gloves when in contact with these patients is greater than when in contact with patients who do not have diarrhea (121). It has also been shown that VRE isolates in the environment surrounding a colonized patient are easily transferred on to the gloved hands of HCWs after contact with environmental surfaces (122). Isolates from patients, environmental surfaces, and gloved hands of HCWs were the same strains by PFGE. Isolates from patients’ intact skin or environmental surfaces may also be transferred to clean sites on patients by HCWs hands or gloves (123). Two studies have shown that environmental surfaces have a lower density of VRE than do perirectal swabs (123,124). Both studies showed that broth amplification was often necessary to recover VRE from environmental surface samples. However, low density of VRE on environmental surfaces did not prevent transfer. Sixty-nine percent of surfaces from which VRE were transferred were positive by broth amplification culture only (123).

16

Mayhall

Another concern about transfer of VRE from environmental surfaces is that the microorganism can survive on inanimate surfaces from seven days to two months (125,126). Further evidence that VRE may survive for a prolonged period on an inanimate surface and then be transferred to a patient is provided by a report on a VRE outbreak in a burn unit (119). After initial control of the outbreak for five weeks, the outbreak recurred from an electrocardiogram (EKG) lead that had not been cleaned since use on the last patient. In the five-week period, during which the outbreak had been cleared, all weekly patient surveillance cultures and 317 environmental cultures were negative for VRE. The VRE cultured from the EKG lead, the prior patient on which the lead had been used and the patient who acquired the VRE from the EKG lead, were shown to be the same strain by PFGE. The time from use of the EKG lead on the first patient to use on the second patient was 38 days. VRE have also been transmitted between patients by electronic thermometers during an outbreak (127). Restriction endonuclease analysis of plasmid DNA indicated that all clinical isolates and isolates from handles of the electronic thermometers were identical. Risk Factors for Acquisition of VRE in ICUs Adult ICUs. Although many published studies have examined risk factors for nosocomial acquisition of VRE, most have not been well designed. When trying to ascertain risk factors for acquisition, it is important to determine the exact time of colonization or infection by VRE, to use controls that are negative for VRE [as opposed to controls positive for vancomycin-susceptible enterococci (VSE)], and to use multivariable statistics to identify independent risk factors. Some studies of risk factors have included ICUs in addition to other areas of the hospital (Table 4), and others have been limited to ICUs (Table 5). Several of the studies included in Tables 4 and 5 have identified a significant relationship between prior administration of an antimicrobial agent and acquisition of VRE. Drugs listed included cephalosporins, metronidazole, vancomycin, carbapenems, ticarcillin–clavulanate, and quinolones. The antibiotic most often identified as a risk factor was vancomycin. In an extensive study of the effects of antimicrobial agents on fecal flora, it was found that antianaerobic antibiotics promoted high-density

Table 4 Risk Factors for Acquisition of Vancomycin-Resistant Enterococci from Studies of Mixed Patient Populations Publications Loeb et al. (128) Byers et al. (129)

Cetinkaya et al. (130)

a

Protective factors.

Risk factors Cephalosporin use Proximity to an unisolated patient History of major trauma Therapy with metronidazole Vancomycin use Gastrointestinal bleedinga Presence of central venous lines Antacid use Mean daily dose of Vicodin1a

Adjusted odds ratio (95% CI) 13.8 (2.5–76.3) 2.04 (1.32–3.14) 9.27 (1.43–60.3) 3.04 (1.05–8.77) 3.2 (1.7–6.0) 0.26 (0.08–0.79) 2.2 (1.04–4.6) 2.9 (1.5–5.6)

p Value 0.01 0.0014 0.020 0.040 0.0003 0.02 0.04 0.002 0.0003

MRSA/VRE Colonization and Infection in the Critical Care Unit

17

Table 5 Risk Factors for Acquisition of Vancomycin-Resistant Enterococci from Studies in Intensive Care Units Publications

Type of ICU

Karanfil et al. (131)

Cardiothoracic surgery

Vancomycin use

Slaughter et al. (132)

Medical

Length of stay in ICU 5 day Enteral feeding Sucralfate Colonization pressure Proportion of days with enteral feeding Proportion of patient days with cephalosporin use Presence of diarrhea Administration of an antacid Enteral feedings

Bonten et al. (53)

Medical

Falk et al. (119)

Burn

Gardiner et al. (133) Padiglione et al. (134)

Medical

Martinez et al. (135)

Warren et al. (136)

a

Multicenter study—mixed ICUs and transplant units Medical

Medical

Risk factors

Adjusted odds ratio (95% CI) Sole predictor in the logistic regression model 0.08 (0.02–0.39)a 6.09 (1.56–23.7) 3.26 (1.09–9.72) 1.032 (1.012–1.052)b 1.009 (1.000–1.017)b

0.002 0.05

1.007 (0.999–1.015)b

0.11

43.9 (5.5–infinity) 24.2 (2.9–infinity)

0.0001 0.002

19 (2.02–177.9)

< 0.05

Renal unit patients Carbapenems Ticarcillin–clavulanate

4.62 (1.22–17)b 2.84 (1.02–7.96)b 3.64 (1.13–11.64)b

0.02 0.048 0.03

Hospitalization for more than one week before MICU admission Administration of vancomycin before or during an ICU admission Administration of quinolones before or during MICU admission Location in a high risk MICU roomc Increasing age Hospitalization in the 6 mo prior to current admission Admission from a long-term care facility

18.5 (1.1–301.0)

0.04

6.3 (1.2–34.0)

0.03

14.8 (1.2–180.0)

0.04

81.7 (2.2–3092.0)

0.02

1.02 (1.01–1.03) 2.74 (2.21–3.40)

1.30 (1.14–1.47)

Protective factor. Hazard ratios. c A room that proved to be contaminated after postpatient discharge cleaning. Abbreviations: ICU, intensive care unit; MICU, medical intensive care unit. b

p Value

18

Mayhall

colonization of stool with VRE (136). Administration of vancomycin had no effect on the concentration of VRE in stool. Although antianaerobic agents increased the concentration of VRE in stool, it is unclear whether these agents or vancomycin predispose to acquisition of VRE. Several case–control studies have shown that vancomycin is a risk factor for acquisition of VRE. In an assessment of studies showing a relationship between vancomycin and acquisition of VRE by meta-analysis, the authors concluded that the apparent relationship between administration of vancomycin and colonization with VRE is due to selection of VSE as the reference group, confounding by duration of hospitalization and publication bias (137). However, several studies have been published in which the reference group was appropriately selected (VRE-negative patients and not VSE-culture positive) and duration of hospitalization was included to control for confounding due to longer exposure time (130,131,138,139). Thus, the issue of whether vancomycin is a risk factor for acquisition of VRE is unsettled. Risk factors from Tables 4 and 5 that appear greater than or equal to two times are use of antacids and enteral feedings. One study noted that a length of stay of less than or equal to five days in an MICU was protective against VRE acquisition, whereas another study observed that hospitalization for more than one week prior to MICU admission was a risk factor for acquisition of VRE. In summary the most frequently identified risk factors for acquiring VRE from these studies are administration of antibiotics and antacids, enteral feedings, and longer length of stay. Neonatal ICUs. There are six reports of outbreaks of VRE in NICUs (113–115,140–142). Analytical epidemiology was used in only one of the studies to identify risk factors for acquisition of VRE (113). This study examined a large number of variables by univariate analysis and found many variables apparently related to VRE colonization. However, multivariable analysis by logistic regression identified days of antimicrobial therapy (OR 1.21, 95% CI 1.045–1.400, p ¼ 0.01) and birth weight (OR 0.92, 95% CI 0.862–0.979, p ¼ 0.009) as the only independent associations with acquisition of VRE. Additional studies are needed to further define the variables associated with acquisition of VRE in this population. Prevention and Control of VRE in ICUs Although less data were available 10 years ago on the epidemiology and control of VRE, recommendations of the CDC’s Hospital Infection Control Practices Advisory Committee (HICPAC) have stood the test of time (143). Virtually all of HICPAC’s recommendations to prevent and control the spread of VRE have been supported by the studies published in the last 10 years. Thus, the focus for control and prevention is on the following: (i) detection of colonized patients by surveillance cultures; (ii) barrier isolation; (iii) hand hygiene; (iv) environmental decontamination; (v) decolonization of HCWs; and (vi) control of antimicrobial (particularly vancomycin) use. The HICPAC guideline also emphasized that prevention and control should start in ICUs and other areas where the VRE transmission rate is the highest. Culture Surveillance Because only about 10% of patients colonized with VRE develop infection, most patients who make up the reservoir of VRE in the hospital are colonized and not infected. Colonization can be detected only by surveillance cultures. Colonized patients have been detected by screening stool specimens submitted to the clinical

MRSA/VRE Colonization and Infection in the Critical Care Unit

19

microbiology laboratory for Clostridium difficile toxin assay (144). Stool may be collected and sent from the ICU to the clinical microbiology laboratory, but in most cases perirectal swab specimens are cultured in broth or streaked to solid agar. One group of authors found that a rectal swab sample had a sensitivity of 58% in detecting VRE compared to culture of stool (145). These authors also noted that the concentration of VRE in stool increased with the number of antibiotics administered and duration of their administration. It is likely that perirectal swab cultures will have a higher sensitivity for detection of VRE in ICUs where many patients are on antibiotics. In another study in a burn unit, the authors observed that perirectal swabs had the same sensitivity for detecting VRE whether inoculated to broth or to solid media (124). This suggests that small numbers of VRE detected by broth amplification can also be detected by growth on solid media. This may have been due to the extensive use of antimicrobial agents in the burn unit where the study was performed. The HICPAC guideline also recommends culturing urine and wounds for VRE (143). This will likely increase the sensitivity of surveillance cultures. Surveillance cultures can be made more efficient by using a selective culture media to suppress growth of other microorganisms that will likely contaminate the specimens (124,143). It is likely that most patients who are colonized with VRE in an ICU will be detected by perirectal swabs and swabs of open wounds and other skin sites inoculated to selective media. This recommendation is further supported by a study that found that rectal and perirectal swabs had approximately the same sensitivity (79%) (146). Surveillance cultures and isolation of colonized and infected patients has been shown in many studies to control VRE in both acute care and long-term care facilities (117,119,120,129,147–150). One publication describes the effective control of VRE in four acute-care hospitals and in 26 long-term care facilities in the Siouxland region of Iowa, Nebraska, and South Dakota (147). Barrier Precautions Patients with VRE infections and VRE colonization detected by surveillance cultures should be immediately placed on barrier or isolation precautions. The HICPAC guideline recommends placement of patients in a single room or in the same room as other patients with VRE (143). The guideline also recommends donning clean nonsterile gloves prior to entering the room. Use of a gown is recommended only for substantial contact with the patient or environment. Many health care facilities now require that a gown be worn as well as gloves to enter the room of a patient with VRE. This is based on several studies. The first report noted that an outbreak in an ICU was not contained until personnel began to wear gowns in addition to gloves (151). In a prospective, controlled nonrandomized study in an MICU in a hospital in which VRE were endemic, half the patients were isolated with glove use alone and the other half were cared for by HCWs wearing both gloves and gowns (132). The authors observed that there was no difference in transmission from patients isolated with glove use only and those isolated with HCWs wearing gowns and gloves. Two prospective nonrandomized studies using historical controls and multivariable analysis carried out in MICUs both observed a significantly lower rate of transmission of VRE when gowns and gloves rather than gloves alone were used for isolation (152,153). In the former study, the addition of gowns was protective only for those patients exposed to VRE for more than 15 days.

20

Mayhall

Use of gowns in addition to gloves is further supported by the findings from a study that evaluated the proportion of gloves, gowns, and stethoscopes that were contaminated after a structured physical examination of patients colonized or infected with VRE (154). Gloves were contaminated in 63%, gowns in 37%, and stethoscopes in 31% of the examinations. Available published data support a recommendation that both gloves and gowns be worn when entering a room where a patient is isolated for VRE colonization or infection. There are few data on when patients colonized or infected with VRE may be taken off of isolation. The CDC’s HICPAC recommendation was that isolation be discontinued when three sets of cultures taken from stool or by rectal swab and all previous positive body sites were culture negative for VRE on three occasions at least one week apart (143). One study has been published that supports the recommendation made by HICPAC that patients may be taken off of isolation after three consecutive negative cultures (155). Decontamination of the Environment That VRE can remain viable on inanimate surfaces from seven days to two months has already been established (119,125,126). In addition to hard surfaces, upholstered surfaces in hospitals can be contaminated with VRE (156). VRE were recovered at 72 hours and one week after inoculation to an upholstered surface. VRE were also recovered from 3 of 10 seat cushions that were cultured in the room of a VRE patient. The authors state that an easily cleanable nonporous material is the preferred upholstery in hospitals. Extensive cultures of environmental surfaces in rooms of patients colonized with VRE in an MICU and a burn ICU identified contaminated surfaces in 12% and 13.5%, respectively (116,119). It has also been shown that at least one environmental surface was positive in the rooms of 63% to 92% of patients colonized with VRE (116,122). Three studies have demonstrated that VRE are easily transferred to gloves or hands of HCWs after contact with the environment (121–123). In one of the latter studies, VRE were transferred from a culture-positive site to a culturenegative site in 10.6% of the opportunities (123). VRE were transferred from patient to environment and from environment to patient. VRE were transferred from sites with low-density contamination or colonization (cultured from broth only) 69% of the time. Environmental contamination has also been shown to be more widely distributed in the areas around the bed of a colonized patient with diarrhea (151). Further evidence for the importance of environmental contamination in the acquisition of VRE in an MICU was the finding in a case–control study that environmental contamination was a risk factor for patients acquiring VRE (135). The effectiveness of decontamination of the environment depends on the method used. In one study, the investigators observed that cleaning environmental surfaces with a cleaning rag sprayed with a quaternary ammonium disinfectant was significantly less effective than dipping the cleaning rag into a bucket of the same disinfectant, drenching all surfaces, allowing the surfaces to remain wet for 10 minutes, and then wiping the surfaces dry with a clean towel (157). The authors referred to the latter method as the bucket method. Using the method in which the disinfectant was sprayed on the cleaning rag took 2.8 applications to eradicate VRE from environmental surfaces compared with one application using the bucket method. In addition to a greater efficiency at removing VRE from surfaces, the bucket method also cost less than the method of spraying disinfectant on a cleaning rag. Based on this study, the bucket method is the preferred method for decontaminating environment surfaces.

MRSA/VRE Colonization and Infection in the Critical Care Unit

21

Hand Hygiene Excellent hand hygiene must always be practiced for the prevention of nosocomial infections, but it is particularly important in providing effective isolation of patients with VRE. Given the frequent contamination of gloved and ungloved hands of HCWs in contact with VRE-colonized patients and environmental surfaces, excellent hand hygiene must be an integral part of barrier precautions for VRE (121–123,154). After patient contact, hands should be washed with an antiseptic-containing soap or an alcohol hand rub should be applied. Colonization of Healthcare Workers Colonization of HCWs with VRE has not been reported in the literature during outbreaks of VRE infection and colonization. A study of 55 stool specimens from HCWs in a hospital where 15% of enterococci were VRE found that all cultures of stool specimens were negative for VRE (158). The authors concluded that colonization resistance was sufficient to prevent colonization of HCW’s gastrointestinal tracts in the absence of acute illness or severe underlying comorbidities. Antimicrobial Agents Antimicrobial agents have been identified as risk factors for acquisition of VRE as shown in Tables 4 and 5. Vancomycin has been considered as a risk factor for acquisition of VRE, but several studies have failed to identify vancomycin as a risk factor (128,129,132,134). The HICPAC recommendations included a list of indications for use of vancomycin and a list of contraindications for use of this antibiotic (143). A more recent publication from the CDC reports on a study performed in cooperation with 20 hospitals in the NNIS system that joined the Intensive Care Antimicrobial Resistance Epidemiology (ICARE) Project. These hospitals contributed data from 50 ICUs on grams of selected antibiotics used each month and on susceptibility tests for selected microorganisms recovered from patients in these units each month (159). The data submitted to Project ICARE was used to create benchmarks for vancomycin use. Those ICUs that instituted changes in practice observed significant decreases in vancomycin use and in VRE prevalence. Although some controversy remains about whether vancomycin use is a risk factor for acquisition of VRE, the bulk of the data to date is in favor of limiting vancomycin use in ICUs as part of the control programs for VRE. Other antibiotics that have been identified as risk factors for acquisition of VRE include cephalosporins, metronidazole, carbapenems, ticarcillin–clavulanate, and quinolones (128,129,134,135). A study of the effect of antimicrobial therapy on the concentration of VRE in patients’ stools observed that concentrations of VRE increased significantly in stools of those patients who received antianaerobic antibiotics. The authors made the point that vancomycin has antianaerobic activity and showed that VRE increased in concentration in stools of patients who were treated with vancomycin (137). The authors also showed that patients with high concentrations of VRE in stools caused greater environmental contamination and observed that eight patients with VRE cultured from blood, urine, and a sacral wound had 6 logs of VRE per gram of stool. Therefore, avoiding the use of antianaerobic antimicrobial therapy in patients when possible may aid in control of VRE by reducing environmental contamination. Limiting the concentration of VRE in stool may also reduce the risk of invasive disease due to VRE. Limiting the use of antianaerobic agents and vancomycin appears important in the control of VRE.

22

Mayhall

Table 6 Control Measures for Vancomycin-Resistant Enterococci in Intensive Care Units Measure Culture patients on admission who are transferred from other healthcare facilities or long-term care facilities and those hospitalized in the last year and weekly while in the ICU until they become positive for VRE or they are discharged Place patients with VRE infection and colonization on contact precautions

Practice hand hygiene after leaving the room

Culture environmental surfaces to assess extent of contamination with VRE

When possible, limit use of those antimicrobial agents that have been identified as risk factors for VRE acquisition or that increase the concentration of VRE in stool

Patients may be taken off of contact precautions when they have had three consecutive sets of negative cultures for VRE, each taken 1 wk apart

Comments Use selective culture media Take specimens for culture from perirectal area and wounds Flag patients’ charts or flag patients in the hospital computer system who are VRE-positive Place patients flagged for VRE on contact precautions on admission Healthcare workers should wear both gown and gloves Masks are not needed Remove gown and gloves prior to leaving the room Wash hands with a soap containing an antiseptic or apply an alcohol hand rub If hands are visibly soiled, wash with a soap containing an antiseptic or wash with plain soap followed by application of an alcohol hand rub Obtain specimens with sterile swabs moistened with sterile saline without bacteriostatic agents Use selective culture media to maximize efficiency of laboratory identification of VRE Use bucket method to clean and disinfect environmental surfaces Culture environmental surfaces to determine the effectiveness of the cleaning and disinfection methods Do not use phenolic disinfectants in NICUs for environmental decontamination Antimicrobial agents that have been identified as risk factors for the acquisition of VRE include cephalosporins (particularly third generation cephalosporins), vancomycin, metronidazole, carbapenems, and ticarcillin–clavulanate Antimicrobial agents that have been shown to increase the concentration of VRE in stools include clindamycin, metronidazole, cefoxitin, and ceftriaxone Piperacillin-tazobactam may protect against acquisition of VRE Cultures should be taken from the perirectal area and all previously positive sites

Abbreviations: VRE, vancomycin-resistant enterococci; ICU, intensive care unit; NICUs, neonatal intensive care units.

MRSA/VRE Colonization and Infection in the Critical Care Unit

23

Another approach to controlling VRE through changes in the use of antimicrobial agents is to replace the use of antimicrobials to which VRE are resistant with antimicrobials to which VRE are more susceptible. Piperacillin/tazobactam has been considered to be a good candidate for suppressing the growth of VRE, because it has good antimicrobial activity against E. faecium, which is the most common VRE species, and because it is concentrated in bile. Five studies on the use of piperacillin–tazobactam in place of third-generation cephalosporins and ticarcillin– clavulanate have been published (160–164). Only one of the latter studies was adequately designed to provide definitive results (164). There was a significant reduction in the acquisition of VRE after ticarcillin–clavulanate was replaced by piperacillin–tazobactam. As the authors pointed out, additional studies are needed for this control strategy as the study was carried out in a single institution and the reduction in acquisition of VRE was associated with the formulary change, but causality could not be established. When other measures have failed to control the spread of VRE, this approach could be tried. In summary when measures are being instituted in an attempt to control VRE, it would appear prudent to limit the use of vancomycin, cephalosporins, metronidazole, clindamycin, and ticarcillin–clavulanate. Initiating the use of piperacillin–tazobactam might add to the effectiveness of manipulating antimicrobials as part of the control measures for VRE. Other risk factors that should be addressed are the use of enteric feedings, the use of antacids, and effectively removing VRE from environmental surfaces. Table 6 lists the control measures for VRE in ICUs. Cost Effectiveness of VRE Control The high cost of VRE control is often mentioned in the literature, and many infection control programs have decided to apply very limited control measures to prevent and control the spread of VRE. However, several recent studies on the cost effectiveness of VRE control have all concluded that effective VRE control with a reduction in infections caused by VRE is cost effective (165–168). In three of the studies, control of VRE was cost effective with savings to the hospitals of between $100,000 and $500,000 per year (165,166,168). The other study estimated the costs of VRE infections in a hospital using a retrospective matched cohort study (167). The authors estimated that the effects of VRE infections on patients would include 15 cases of in-hospital deaths, 22 major operations, 26 ICU admissions, and 1445 additional hospitalization days with excess costs of $2,974,478 during the study period. It is reasonable to conclude from the available data that control of VRE is cost effective. REFERENCES 1. Jevons MP. ‘‘Celbenin’’– resistant staphylococci. Br Med J 1961; 1:124–125. 2. Barrett FF, McGehee RF, Finland M. Methicillin-resistant Staphylococcus aureus at Boston City Hospital. N Engl J Med 1968; 279:441–448. 3. http://www.cdc.gov/ncidod/hip/Aresist/ICU-RES Trend 1995–2004.pdf (accessed June2005). 4. Daum RS, Ito T, Hiramatsu K, et al. A novel methicillin-resistance cassette in community-acquired methicillin-resistant Staphylococcus aureus isolates of diverse genetic backgrounds. J Infect Dis 2002; 186(9):1344–1347.

24

Mayhall

5. Salmenlinna S, Lyytika¨inen O, Vuopio-Varkila J. Community-acquired methicillinresistant Staphylococcus aureus, Finland. Emerg Infect Dis 2002; 8(6):602–607. ¨ sterlund A, Kahlmeter G, Bieber L, et al. Intrafamilial spread of highly virulent 6. O Staphylococcus aureus strains carrying the gene for Panton-Valentine leukocidin. Scand J Infect Dis 2002; 34:763–764. 7. Liassine N, Auckenthaler R, Descombes M-C, et al. Community-acquired methicillinresistant Staphylococcus aureus isolated in Switzerland contains the Panton-Valentine leukocidin or exfoliative toxin genes. J Clin Microbiol 2004; 42(2):825–828. 8. Stemper ME, Shukla SK, Reed KD. Emergence and spread of community-associated methicillin-resistant Staphylococcus aureus in rural Wisconsin, 1989 to 1999. J Clin Microbiol 2004; 42(12):5673–5680. 9. Diep BA, Sensabaugh GF, Somboona NS, et al. Widespread skin and soft-tissue infections due to two methicillin-resistant Staphylococcus aureus strains harboring the genes for Panton-Valentine Leucocidin. J Clin Microbiol 2004; 42(5):2080–2084. 10. Cooms GW, Nimmo GR, Bell JM, et al. Genetic diversity among community methicillinresistant Staphylococcus aureus strains causing outpatient infections in Australia. J Clin Microbiol 2004; 42(10):4735–4743. 11. Witte W, Braulke C, Cuny C, et al. Emergence of methicillin-resistant Staphylococcus aureus with Panton-Valentine leukocidin genes in central Europe. Eur J Clin Microbiol Infect Dis 2005; 24:1–5. 12. Lu P-L, Chin L-C, Peng C-F, et al. Risk factors and molecular analysis of community methicillin-resistant Staphylococcus aureus carriage. J Clin Microbiol 2005; 43(1):132–139. 13. Hussain FM, Boyle-Vavra S, Daum RS. Community-acquired methicillin-resistant Staphylococcus aureus colonization in healthy children attending an outpatient pediatric clinic. Pediatr Infect Dis J 2001; 20(8):763–767. 14. Naimi TS, LeDell KH, Boxrud DJ, et al. Epidemiology and clonality of communityacquired methicillin-resistant Staphylococcus aureus in Minnesota 1996–1998. Clin Infect Dis 2001; 33(7):990–996. 15. Fridkin SK, Hageman JC, Morrison M, et al. Methicillin-resistant Staphylococcus aureus disease in three communities. N Engl J Med 2005; 352(14):1436–1444. 16. Morel A-S, Wu F, Della-Latta P, et al. Nosocomial transmission of methicillin-resistant Staphylococcus aureus from a mother to her preterm quadruplet infants. Am J Infect Control 2002; 30(3):170–173. 17. Regev-Yochay G, Rubinstein E, Barzilai A, et al. Methicillin-resistant Staphylococcus aureus in neonatal intensive care unit. Emerg Infect Dis 2005; 11(3):453–456. 18. Uttley AH, Collins CH, Naidoo J, et al. Vancomycin-resistant enterococci. Lancet 1988; 1(1):57–58. 19. Leclercq R, Derlot E, Duval J, et al. Plasmid-mediated resistance to vancomycin and teicoplanin in Enterococcus faecium. N Engl J Med 1998; 319(3):157–161. 20. Nosocomial enterococci resistant to vancomycin-United States, 1989–1993. Morbid Mortal Wkly Rep 1993; 42(30):597–600. 21. Centers for Disease Control and Prevention. National Nosocomial Infections Surveillance (NNIS) system report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004; 32(12):470–485. 22. Montecalvo MA, de Lencastre H, Carraher M, et al. Natural history of colonization with vancomycin-resistant Enterococcus faecium. Infect Control Hosp Epidemiol 1995; 16(12):680–685. 23. Henning KJ, de Lencastre H, Eagan, J, et al. Vancomycin-resistant Enterococcus faecium on a pediatric oncology ward: duration of stool shedding and incidence of clinical infection. Pediatr Infect Dis J 1996; 15(10):848–854. 24. Patel R. Clinical impact of vancomycin-resistant enterococci. J Antimicrob Chemother 2003; 51(suppl 3):iii13–iii21. 25. McGeer AJ, Low DE. Vancomycin-resistant enterococci. Semin Respir Infect 2000; 15(4):314–326.

MRSA/VRE Colonization and Infection in the Critical Care Unit

25

26. Murray BE. Drug therapy: vancomycin-resistant enterococcal infections. N Engl J Med 2000; 342(10):710–721. 27. Kreiswirth B, Kornblum J, Arbeit RD, et al. Evidence for a clonal origin of methicillin resistance in Staphylococcus aureus. Science 1993; 259(1):227–230. 28. Musser JM, Kapur V. Clonal anaylsis of methicillin-resistant Staphylococcus aureus strains from intercontinental sources: association of the mec gene with divergent phylogenetic lineages implies dissemination by horizontal transfer and recombination. J Clin Microbiol 1992; 30:2058–2063. 29. Mongkolrattanothai K, Boyle S, Kahana MD, et al. Severe Staphylococcus aureus infections caused by clonally related community-acquired methicillin-susceptible and methicillin-resistant isolates. Clin Infect Dis 2003; 37(8):1050–1058. 30. Eady EA, Cove JH. Staphylococcal resistance revisited: community-acquired methicillin resistant Staphylococcus aureus – an emerging problem for the management of skin and soft tissue infections. Curr Opin Infect Dis 2003; 16:103–124. 31. Charlebois ED, Perdreau-Remington F, Kreiswirth B, et al. Origins of community strains of methicillin-resistant Staphylococcus aureus. Clin Infect Dis 2004; 39(1):47–54. 32. Baba T, Takenchi F, Kuroda M, et al. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet 2002; 359:1819–1827. 33. Eguia JM, Chambers HF. Community-acquired methicillin-resistant Staphylococcus aureus: epidemiology and potential virulence factors. Curr Infect Dis Rep 2003; 5: 459–466. 34. Chaix C, Durand-Zaleski I, Alberti C, et al. Control of endemic methicillin-resistant Staphylococcus aureus: a cost-benefit analysis in an intensive care unit. JAMA 1999; 282(18):1745–1751. 35. Squier C, Rihs JD, Risa J, et al. Staphylococcus aureus rectal carriage and its association with infections in patients in a surgical intensive care unit and a liver transplant unit. Infect Control Hosp Epidemiol 2002; 23(9):495–501. 36. Marshall C, Harrington G, Wolfe R, et al. Acquisition of methicillin-resistant Staphylococcus aureus in a large intensive care unit. Infect Control Hosp Epidemiol 2003; 24(5):322–326. 37. Cassone M, Campanile F, Pantosti A, et al. Identification of a variant ‘‘Rome clone’’ of methicillin-resistant Staphylococcus aureus with decreased susceptibility to vancomycin, responsible for an outbreak in an intensive care unit. Microb Drug Resist 2004; 10:43–49. 38. Thompson DS. Methicillin-resistant Staphylococcus aureus in a general intensive care unit. JR Soc Med 2004; 97:521–526. 39. Gastmeier P, Sohr D, Geffers C, et al. Mortality risk factors with noscomial Staphylococcus aureus infections in intensive care units: results from the German Nosocomial Infection Surveillance System (KISS). Infection 2005; 33(2):50–55. 40. Haley RW, Cushion NB, Tenover FC, et al. Eradication of endemic methicillin-resistant Staphylococcus aureus infections from a neonatal intensive care unit. J Infect Dis 1995; 171(3):614–624. 41. Andersen BM, Lindermann R, Bergh K, et al. Spread of methicillin-resistant Staphylococcus aureus in a neonatal intensive unit associated with understaffing, overcrowding and mixing of patients. J Hosp Infect 2002; 50(1):18–24. 42. Nambiar S, Herwaldt LA, Singh N. Outbreak of invasive disease caused by methicillinresistant Staphylococcus aureus in neonates and prevalence in the neonatal intensive care unit. Pediatr Crit Care Med 2003; 4(2):220–226. 43. Isaacs D, Fraser S, Hogg G, et al. Staphylococcus aureus infections in Australasian neonatal nurseries. Arch Dis Child Fetal Neonatal Ed 2004; 89:F331–F335. 44. Saiman L, Cronquist A, Wu F, et al. An outbreak of methicillin-resistant Staphylococcus aureus in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2003; 24(5): 317–321. 45. Udo EE, Pearman JW, Grubb WB. Genetic analysis of community isolates of methicillin-resistant Staphylococcus aureus in Western Australia. J Hosp Infect 1993; 25:97–108.

26

Mayhall

46. Maguire GP, Arthur AD, Boustead PJ, et al. Clinical experience and outcomes of community-acquired and nosocomial methicillin-resistant Staphylococcus aureus in a northern Australian hospital. J Hosp Infect 1998; 38:273–281. 47. O’Brien FG, Pearman JW, Gracy M, et al. Community strain of methicillin-resistant Staphylococcus aureus involved in a hospital outbreak. J Clin Microbiol 1999; 37(9): 2858–2862. 48. Eckhardt C, Halvosa JS, Ray SM, et al. Transmission of methicillin-resistant Staphylococcus aureus in the neonatal intensive care unit from a patient with community-acquired disease. Infect Control Hospital Epidemiol 2003; 24(6):460–461. 49. Bratu S, Eramo A, Kopec R, et al. Community-associated methicillin-resistant Staphylococcus aureus in hospital nursery and maternity units. Emerg Infect Dis 2005; 11(6): 808–813. 50. Merrer J, Santoli F, Appe´re´-De Vecchi C, et al. ‘‘Colonization pressure’’ and risk of acquisition of methicillin-resistant Staphylococcus aureus in a medical intensive care unit. Infect Control Hosp Epidemiol 2000; 21(11):718–723. 51. Troche´ G, Joly L-M, Guibert M, et al. Detection and treatment of antibiotic-resistant bacterial carriage in a surgical intensive care unit: a 6-year prospective survey. Infect Control Hosp Epidemiol 2005; 26(2):161–165. 52. Grundmann H, Hori S, Winter B, et al. Risk factors for the transmission of methicillinresistant Staphylococcus aureus in an adult intensive care unit: fitting a model to the data. J Infect Dis 2002; 185(4):481–488. 53. Bonten MJM, Slaughter S, Ambergen AW, et al. The role of ‘‘colonization pressure’’ in the spread of vancomycin-resistant enterocci. An important infection control variable. Arch Intern Med 1998; 158:1127–1132. 54. Simor AE, Ofner-Agostini M, Bryce E, et al. The evolution of methicillin-resistant Staphylococcus aureus in Canadian hospitals: 5 years of national surveillance. CMAJ 2001; 165(1):21–26. 55. Dupeyron C, Campillo B, Bordes M, et al. A clinical trial of mupirocin in the eradication of methicillin-resistant Staphylococcus aureus nasal carriage in a digestive disease unit. J Hosp Infect 2000; 52(4):281–287. 56. Boyce JM, Opal SM, Potter-Bynoe G, et al. Spread of methicillin-resistant Staphylococcus aureus in a hospital after exposure to a healthcare worker with chronic sinusitis. Clin Infect Dis 1993; 17(3):496–504. 57. Wang J-T, Chang S-C, Ko W-J, et al. A hospital-acquired outbreak of methicillinresistant Staphylococcus aureus infection initiated by a surgeon carrier. J Hosp Infect 2001; 47(2):104–109. 58. Devine J, Cooke RPD, Wright EP. Is methicillin-resistant Staphylococcus aureus (MRSA) contamination of ward-based computer terminals a surrogate marker for nosocomial MRSA transmission and handwashing compliance? J Hosp Infect 2001; 48(1):72–75. 59. Boyce JM, Potter-Bynoe G, Chenevert C, et al. Environmental contamination due to methicillin-resistant Staphylococcus aureus: possible infection control implications. Infect Control Hosp Epidemiol 1997; 18(9):622–627. 60. Wagenvoort JHT, Sluijsmans W, Penders RJR. Better environmental survival of outbreak versus sporadic MRSA isolates. J Hosp Infect 2000; 45(3):231–234. 61. Bhalla A, Pultz NJ, Gries DM, et al. Acquisition of nosocomial pathogens on hands after contact with environmental surfaces near hospitalized patients. Infect Control Hosp Epidemiol 2004; 25(2):164–167. 62. McBryde ES, Bradley LC, Whitby M, et al. An investigation of contact transmission of methicillin-resistant Staphylococcus aureus. J Hosp Infect 2004; 58(2):104–108. 63. Vriens MR, Fluit AC, Troelstra A, et al. Is methicillin-resistant Staphylococcus aureus more contagious than methicillin-susceptible S. aureus in a surgical intensive care unit? Infect Control Hosp Epidemiol 2002; 23(9):491–494. 64. http://www.cdc.gov/ncidod/hip/ARESIST/mrsahcw.htm (accessed July, 2005).

MRSA/VRE Colonization and Infection in the Critical Care Unit

27

65. Shiomori T, Miyamoto H, Makishima K. Significance of airborne transmission of methicillin-resistant Staphylococcus aureus in an otolaryngology-head and neck surgery unit. Arch Otolaryngol Head Neck Surg 2001; 127(6):644–648. 66. Sherertz RJ, Bassetti S, Bassetti-Wyss B. ‘‘Cloud’’ health-care workers. Emerg Infect Dis 2001; 7(2):241–244. 67. Bassetti S, Bischoff WE, Walter M, et al. Dispersal of Staphylococcus aureus into the air associated with a rhinovirus infection. Infect Control Hosp Epidemiol 2005; 26(2):196–203. 68. Marshall C, Wolfe R, Kossmann T, et al. Risk factors for acquisition of methicillinresistant Staphlococcus aureus (MRSA) by trauma patients in the intensive care unit. J Hosp Infect 2004; 57(3):245–252. 69. Healy CM, Hulten KG, Palazzi DL, et al. Emergence of new strains of methicillinresistant Staphylococcus aureus in a neonatal intensive care unit. Clin Infect Dis 2004; 39:1460–1466. 70. Muto CA, Jernigan JA, Ostrowsky BE, et al. SHEA Guideline for preventing nosocomial transmission of multidrug-resistant strains of Staphylococcus aureus and Enterococcus. Infect Control Hosp Epidemiol 2003; 24(5):362–386. 71. Boyce JM, Havill NL, Kohan C, et al. Do infection control measures work for methicillin-resistant Staphylococcus aureus? Infect Control Hosp Epidemiol 2004; 25(5):395–401. 72. Jernigan JA, Titus MG, Groschel DHM, et al. Effectiveness of contact isolation during a hospital outbreak of methicillin-resistant Staphylococcus aureus. Am J Epidemiol 1996; 143:496–504. 73. Salmenlinna S. Lyytika¨inen O, Kotilainen P, et al. Molecular epidemiology of methicillin-resistant Staphylococcus aureus in Finland. Eur J Clin Microbiol Infect Dis 2000; 19:101–107. 74. Thompson RL, Cabezudo I, Wenzel RP. Epidemiology of nosocomial infections caused by methicillin-resistant Staphylococcus aureus. Ann Intern Med 1982; 97:309–317. 75. Jernigan JA, Clemence MA, Stott GA, et al. Control of methicillin-resistant Staphylococcus aureus at a university hospital: one decade later. Infect Control Hosp Epidemiol 1995; 16(12):686–696. 76. Jans B, Suentens C, Struelens M. Decreasing MRSA rates in Belgian hospitals: results from the national surveillance network after introduction of national guidelines. Infect Control Hosp Epidemiol 2000; 21(6):419. 77. Harbarth S, Martin Y, Rohner P, et al. Effect of delayed infection control measures on a hospital outbreak of methicillin-resistant Staphylococcus aureus. J Hosp Infect 2000; 46(1):42–49. 78. Back NA, Linnemann CC Jr., Staneck JL, et al. Control of methicillin-resistant Staphylococcus aureus in a neonatal intensive-care unit: use of intensive microbiologic surveillance and mupirocin. Infect Control Hosp Epidemiol 1996; 17(4):227–231. 79. Nicolle LE, Dyck B, Thompson G, et al. Regional dissemination and control of epidemic methicillin-resistant Staphylococcus aureus. Infect Control Hosp Epidemiol 1999; 20(3):202–205. 80. Law MR, Gillon. Hospital-acquired infection with methicillin-resistant and methicillinsensitive staphylococci. Epidemiol Infect 1988; 101:623–629. 81. Murray-Leisure KA, Gei S, et al. Control of epidemic methicillin-resistant Staphylococcus aureus. Infect Control Hosp Epidemiol 1990; 11:343–350. 82. Kotilainen P, Routamaa M, Peltonen R, et al. Eradication of methicillin-resistant Staphylococcus aureus from a health center ward and associated nursing home. Arch Intern Med 2001; 161:859–863. 83. Lucet J-C, Chevret S, Durand-Zaleski I, et al. Prevalence and risk factors for carriage of methicillin-resistant Staphylococcus aureus at admission to the intensive care unit. Arch Intern Med 2003; 163:181–188. 84. Girou E, Pujade G, Legrand P, et al. Selective screening of carriers for control of methicillin-resistant Staphylococcus aureus (MRSA) in high-risk hospital areas with a high level of endemic MRSA. Clin Infect Dis 1998; 27(3):543–550.

28

Mayhall

85. Francois P, Pittet D, Bento M, et al. Rapid detection of methicillin-resistant Staphylococcus aureus directly from sterile or nonsterile clinical samples by a new molecular assay. J Clin Microbiol 2003; 41:254–260. 86. Boyce JM, Chenevert C. Abstracts of Papers, Eighth Annual Meeting of the Society for Healthcare Epidemiology of America, Orlando, FL, Apr 5–7, 1998. 87. Dietze B, Rath A, Wendt C, et al. Survival of MRSA on sterile goods packaging. J Hosp Infect 2001; 49(4):255–261. 88. Embil J, McLeod J, Al-Barrak AM, et al. An outbreak of the methicillin-resistant Staphylococcus aureus on a burn unit: potential role of contaminated hydrotherapy equipment. Burn 2001; 27:681–688. 89. Rampling A, Wiseman S, Davis L, et al. Evidence that hospital hygiene is important in the control of methicillin resistant Staphylococcus aureus. J Hosp Infect 2001; 49: 109–116. 90. Farr BM. Prevention and control of methicillin-resistant Staphylococcus aureus infections. Curr Opin Infect Dis 2004; 17:317–322. 91. Arnold MS, Dempsey JM, Fishman M, et al. The best hospital practices for controlling methicillin-resistant Staphylococcus aureus: on the cutting edge. Infect Control Hosp Epidemiol 2002; 23(2):69–76. 92. Boyce JM. MRSA patients: proven methods to treat colonization and infection. J Hosp Infect 2001; 48(suppl A):S9–S14. 93. Watanakunakorn C, Axelson C, Bota B, et al. Mupirocin ointment with and without chlorhexidine baths in the eradication of Staphylococcus aureus nasal carriage in nursing home residents. Am J Infect Control 1995; 23:306–309. 94. Vivoni AM, Santos KRN, de-Oliveira MP, et al. Mupirocin for controlling methicillinresistant Staphylococcus aureus: lessons from a decade of use at a university hospital. Infect Control Hosp Epidemiol 2005; 26(7):662–667. 95. Silvestri L, Milanese M, Oblach L, et al. Enteral vancomycin to control methicillinresistant Staphylococcus aureus outbreak in mechanically ventilated patients. Am J Infect Control 2002; 30(7):391–399. 96. Khoury J, Jones M, Grim A, et al. Eradication of methicillin-resistant Staphylococcus aureus from a neonatal intensive care unit by active surveillance and aggressive infection control measures. Infect Control Hosp Epidemiol 2005; 26(7):616–621. 97. Doebbeling BN, Breneman DL, Neu HC, et al. Elimination of Staphylococcus aureus nasal carriage in healthcare workers: analysis of six clinical trials with calcium mupirocin ointment. Clin Infect Dis 1993; 17(9):466–474. 98. Reagan DR, Doebbeling BN, Pfaller MA, et al. Elimination of coincident Staphylococcus aureus nasal and hand carriage with intranasal application of mupirocin calcium ointment. Ann Intern Med 1991; 114(2):101–106. 99. Chenoweth CE. Enterococcus species. In: Mayhall CG, ed. Hospital Epidemiology and Infection Control. 3rd ed. Philadelphia: Lippincott Williams and Wilkins, 2004:529–544. 100. Bugg TDH, Dutka-Malen S, Arthur M, et al. Identification of vancomycin resistance protein VanA as a D-Alanine: D-Alanine ligase of altered substrate specificity. Biochemistry 1991; 30(8):2017–2021. 101. Willems RJL, Homan W, Top J, et al. Variant esp gene as a marker of a distinct genetic lineage of vancomycin-resistant Enterococcus faecium spreading in hospitals. Lancet 2001; 357:853–855. 102. Rice LB, Carias L, Rudin S, et al. A potential virulence gene, hylEfm, predominates in Enterococcus faecium of clinical origin. J Infect Dis 2003; 187(3):508–512. 103. Shay DK, Maloney SA, Monetcalvo M, et al. Epidemiology and mortality risk of vancomycin-resistant enterococcal bloodstream infections. J Infect Dis 1995; 172(10): 993–1000. 104. Montecalvo MA, Shay DK, Patel P, et al. Bloodstream infections with vancomycinresistant enterococci. Arch Intern Med 1996; 156:1458–1462.

MRSA/VRE Colonization and Infection in the Critical Care Unit

29

105. Stroud L, Edwards J, Danzig L, et al. Risk factors for mortality associated with enterococcal bloodstream infections. Infect Control Hosp Epidemiol 1996; 17(9):576–580. 106. Edmond MB, Ober JF, Dawson JD, et al. Vancomycin-resistant enterococcal bacteremia: natural history and attributable mortality. Clin Infect Dis 1996; 23(12):1234–1239. 107. Stosor V, Peterson LR, Postelnick M, et al. Enterococcus faecium bacteremia. Does vancomycin resistance make a difference? Arch Intern Med 1998; 158:522–527. 108. Zaas AK, Song X, Tucker P, et al. Risk factors for development of vancomycin-resistant enterococcal bloodstream infection in patients with cancer who are colonized with vancomycin-resistant enterococci. Clin Infect Dis 2002; 35(10):1139–1146. 109. Raad II, Hanna HA, Boktour M, et al. Catheter-related vancomycin-resistant Enterococcus faecium bacteremia: clinical and molecular epidemiology. Infect Control Hosp Epidemiol 2005; 26(7):658–661. 110. Hussain Z, Shaikh A, Peloquin CA, et al. Successful treatment of vancomycin-resistant Enterococcus faecium meningitis with linezolid: case report and literature review. Scan J Infect Dis 2001; 33:375–379. 111. Steinmetz MP, Vogelbaum MA, DeGeorgia MA, et al. Successful treatment of vancomycin-resistant enterococcus meningitis with linezolid: case report and review of the literature. Crit Care Med 2001; 29(12):2383–2385. 112. Wong AHM, Wenzel RP, Edmond MB. Epidemiology of bacteriuria caused by vancomycin-resistant enterococci: a retrospective study. Am J Infect Control 2000; 28(4):277–281. 113. Malik RK, Montecalvo MA, Reale MR, et al. Epidemiology and control of vancomycinresistant enterococci in a regional neonatal intensive care unit. Pediatr Infect Dis J 1999; 18(4):352–356. 114. Singh N, Leger M-M, Campbell J, et al. Control of vancomycin-resistant enterococci in the neonatal intensive care unit. Infect Control Hosp Epidemiol 2005; 26(7):646–649. 115. Sherer CR, Sprague BM, Campos JM, et al. Characterizing vancomycin-resistant enterococci in neonatal intensive care. Emerg Infect Dis 2005; 11(9):1470–1472. 116. Bonten MJM, Hayden MK, Nathan C, et al. Epidemiology of colonization of patients and environmental with vancomycin-resistant enterococci. Lancet 1996; 348:1615–1619. 117. Montecalvo MA, Shay DK, Gedris C, et al. A semiquantitative analysis of the fecal flora of patients with vancomycin-resistant enterococci: colonized patients pose an infection control risk. Clin Infect Dis 1997; 25(10):929–930. 118. Ostrowsky BE, Venkataraman L, D’Agata EMC, et al. Vancomycin-resistant enterococci in intensive care units: high frequency of stool carriage during a non-outbreak period. Arch Intern Med 1999; 159(13):1467–1472. 119. Falk PS, Winnike J, Woodmansee C, et al. Outbreak of vancomycin-resistant enterococci in a burn unit. Infect Control Hosp Epidemiol 2000; 21(9):575–582. 120. Boyce JM, Mermel LA, Zervos MJ, et al. Controlling vancomycin-resistant enterococci. Infect Control Hosp Epidemiol 1995; 16(11):634–637. 121. Tenorio AR, Badri SM, Sahgal NB, et al. Effectiveness of gloves in the prevention of hand carriage of vancomycin-resistant Enterococcus species by health care workers after patient care. Clin Infect Dis 2001; 32(5):826–829. 122. Ray AJ, Taub TF. Nosocomial transmission of vancomycin-resistant enterococci from surfaces. JAMA 2002; 287(11):1400–1401. 123. Duckro AN, Blom DW, Lyle EA, et al. Transfer of vancomycin-resistant enterococci via health care worker hands. Arch Intern Med 2005; 165:302–307. 124. Reisner BS, Shaw S, Huber ME, et al. Comparison of three methods to recover vancomycin-resistant enterococci (VRE) from perianal and environmental samples collected during a hospital outbreak of VRE. Infect Control Hosp Epidemiol 2000; 21(12):775–779. 125. Noskin GA, Stosor V, Cooper I, et al. Recovery of vancomycin-resistant enterococci on fingertips and environmental surfaces. Infect Control Hosp Epidemiol 1995; 16(10):577–581.

30

Mayhall

126. Bonilla HF, Zervos MJ, Kauffman CA. Long-term survival of vancomycin-resistant Enterococcus faecium on a contaminated surface. Infect Control Hosp Epidemiol 1996; 17(12):770–771. 127. Livornese LL Jr., Dias S, Samel C, et al. Hospital-acquired infection with vancomycinresistant Enterococcus faecium transmitted by electronic thermometers. Ann Intern Med 1992; 117(2):112–116. 128. Loeb M, Salama S, Armstrong-Evans M, et al. A case-control study to detect modifiable risk factors for colonization with vancomycin-resistant enterococci. Infect Control Hosp Epidemiol 1999; 20(11):760–763. 129. Byers KE, Anglim AM, Anneski CJ, et al. A hospital epidemic of vancomycin-resistant Enterococcus: risk factors and control. Infect Control Hosp Epidemiol 2001; 22(3): 140–147. 130. Cetinkaya Y, Falk PS, Mayhall CG. Effect of gastrointestinal bleeding and oral medications on acquisition of vancomycin-resistant Enterococcus faecium in hospitalized patients. Clin Infect Dis 2002; 35(8):935–942. 131. Karanfil LV, Murphy M, Josephson A, et al. A cluster of vancomycin-resistant Enterococcus faecium in an intensive care unit. Infect Control Hosp Epidemiol 1992; 13(4):195–200. 132. Slaughter S, Hayden MK, Nathan C, et al. A comparison of the effect of universal use of gloves and gowns with that of glove use alone on acquisition of vancomycin-resistant enterococci in a medical intensive care unit. Ann Intern Med 1996; 125(6):448–456. 133. Gardiner D, Murphey S, Ossman E, et al. Prevalence and acquisition of vancomycinresistant enterococci in a medical intensive care unit. Infect Control Hosp Epidemiol 2002; 23(8):466–468. 134. Padiglione AA, Wolfe R, Grabsch EA, et al. Risk factors for new detection of vancomycin-resistant enterococci in acute-care hospitals that employ strict infection control procedures. Antimicrob Agents Chemother 2003; 47(8):2492–2498. 135. Martinez JA, Ruthazer R, Hansjosten K, et al. Role of environmental contamination as a risk factor for acquisition of vancomycin-resistant enterococci in patients treated in a medical intensive care unit. Arch Intern Med 2003; 163:1905–1912. 136. Warren DK, Nitin A, Hill C, et al. Occurrence of co-colonization or co-infection with vancomycin-resistant enterococci and methicillin-resistant Staphylococcus aureus in a medical intensive care unit. Infect Control Hosp Epidemiol 2004; 25(2):99–104. 137. Donskey CJ, Chowdhry TK, Hecker MT, et al. Effect of antibiotic therapy on the density of vancomycin-resistant enterococci in the stool of colonized patients. N Engl J Med 2000; 343(26):1925–1932. 138. Carmeli Y, Samore MH, Huskins WC. The association between antecedent vancomycin treatment and hospital-acquired vancomycin-resistant enterococci. A meta-analysis. Arch Intern Med 1999; 159:2461–2468. 139. D’Agata EMC, Green WK, Schulman G, et al. Vancomycin-resistant enterococci among chronic hemodialysis patients: a prospective study of acquisition. Clin Infect Dis 2001; 32(1):23–29. 140. Lee HK, Lee WG, Cho SR. Clinical and molecular biological analysis of a nosocomial outbreak of vancomycin-resistant enterococci in a neonatal intensive care unit. Acta Pediatr 1999; 88:651–654. 141. Yu¨ce A, Karaman M, Gu¨lay Z, et al. Vancomycin-resistant enterococci in neonates. Scand J Infect Dis 2001; 33:803–805. 142. Rupp ME, Marion N, Fey PD, et al. Outbreak of vancomycin-resistant Enterococcus faecium in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2001; 22(5):301–303. 143. Recommendations for preventing the spread of vancomycin resistance. Hospital Infection Control Practices Advisory Committee (HICPAC). Infect Control Hosp Epidemiol 1995; 16(2):105–113.

MRSA/VRE Colonization and Infection in the Critical Care Unit

31

144. Leber AL, Hindler JF, Kato EO, et al. Laboratory-based surveillance for vancomycinresistant enterococci: utility of screening stool specimens submitted for Clostridium difficile toxin assay. Infect Control Hosp Epidemiol 2001; 22(3):160–164. 145. D’Agata EMC, Gautam S, Green WK, et al. High rate of false-negative results of the rectal swab culture method in detection of gastrointestinal colonization with vancomycinresistant enterococci. Clin Infect Dis 2002; 34(2):167–172. 146. Weinstein JW, Tallapragada S, Farrel P, et al. Comparison of rectal and perirectal swabs for detection of colonization with vancomycin-resistant enterococci. J Clin Microbiol 1996; 34(1):210–212. 147. Ostrowsky BE, Trick WE, Sohn AH, et al. Control of vancomycin-resistant enterococcus in health care facilities in a region. N Engl J Med 2001; 344(19):1427–1433. 148. Siddiqui AH, Harris AD, Hebden J, et al. The effect of active surveillance for vancomycin-resistant enterococci in high-risk units on vancomycin-resistant enterococci incidence hospital-wide. Am J Infect Control 2002; 30(1):40–43. 149. Calfee DP, Giannetta ET, Durbin LJ, et al. Control of endemic vancomycin-resistant Enterococcus among inpatients at a university hospital. Clin Infect Dis 2003; 37(3): 326–332. 150. Price CS, Paule S, Noskin GA, et al. Active surveillance reduces the incidence of vancomycin-resistant enterococcal bacteremia. Clin Infect Dis 2003; 37(7):921–928. 151. Boyce JM, Opal SM, Chow JW, et al. Outbreak of multidrug-resistant Enterococcus faecium with transferable vanB class vancomycin resistance. J Clin Microbiol 1994; 32(5):1148–1153. 152. Puzniak LA, Leet T, Mayfield J, et al. To gown or not to gown: the effect on acquisition of vancomycin-resistant enterococci. Clin Infect Dis 2002; 35(1):18–25. 153. Srinivasan A, Song X, Ross T, et al. A prospective study to determine whether cover gowns in addition to gloves decrease nosocomial transmission of vancomycin-resistant enterococci in an intensive care unit. Infect Control Hosp Epidemiol 2002; 23(8):424–428. 154. Zachary KC, Bayne PS, Morrison VJ, et al. Contamination of gowns, gloves and stethoscopes with vancomycin-resistant enterococci. Infect Control Hosp Epidemiol 2001; 22(9):560–564. 155. Byers KE, Anglim AM, Anneski CJ, et al. Duration of colonization with vancomycinresistant Enterococcus. Infect Control Hosp Epidemiol 2002; 23(4):207–211. 156. Noskin GA, Bednarz P, Suriano T, et al. Persistent contamination of fabric covered furniture by vancomycin-resistant enterococci: implications for upholstery selection in hospitals. Am J Infect Control 2000; 28(4):311–313. 157. Byers KE, Durbin LJ, Simonton BM, et al. Disinfection of hospital rooms contaminated with vancomycin-resistant Enterococcu faecium. Infect Control Hsop Epidemiol 1998; 19(4):261–264. 158. Carmeli Y, Venkataraman L, DeGirolami PC, et al. Stool colonization of healthcare workers with selected resistant bacteria. Infect Control Hosp Epidemiol 1998; 19(1):38–40. 159. Fridkin SK, Lawton R, Edwards JR, et al. Monitoring antimicrobial use and resistance: comparison with a national benchmark on reducing vancomycin use and vancomycinresistant enterococci. Emerg Infect Dis 2002; 8(7):702–707. 160. Quale J, Landman D, Saurina G, et al. Manipulation of a hospital antimicrobial formulary to control an outbreak of vancomycin-resistant enterococci. Clin Infect Dis 1996; 23(11):1020–1025. 161. May AK, Melton SM, McGwin G, et al. Reduction of vancomycin-resistant enterococcal infections by limitation of broad-spectrum cephalosporin use in a trauma and burn intensive care unit. Shock 2000; 14(3):259–264. 162. Chavers LS, Moser SA, Funkhouser E, et al. Association between antecedent intravenous antimicrobial exposure and isolation of vancomycin-resistant enterococci. Microb Drug Resist 2003; 9(suppl 1):S69–S77.

32

Mayhall

163. Stiefel U, Paterson DL, Pultz NJ, et al. Effect of the increasing use of piperacillin/ tazobactam on the incidence of vancomycin-resistant enterococci in four academic medical centers. Infect Control Hosp Epidemiol 2004; 25(5):380–383. 164. Winston LG, Charlebois ED, Pang S, et al. Impact of a formulary switch from ticarcillinclavulanate to piperacillin-tazobactam on colonization with vancomycin-resistant enterococci. Am J Infect Control 2004; 32(8):462–469. 165. Montecalvo MA, Jarvis WR, Uman J, et al. Costs and savings associated with infection control measures that reduced transmission of vancomycin-resistant enterococci in an endemic setting. Infect Control Hosp Epidemiol 2001; 22(7):437–442. 166. Muto CA, Giannetta ET, Durbin LJ, et al. Cost-effectiveness of perirectal surveillance cultures for controlling vancomycin-resistant Enterococcus. Infect Control Hosp Epidemiol 2002; 23(8):429–435. 167. Carmeli Y, Eliopoulos G, Mozaffari E, et al. Health and economic outcomes of vancomycin-resistant enterococci. Arch Intern Med 2002; 162:2223–2228. 168. Puzniak LA, Gillespie KN, Leet T, et al. A cost-benefit analysis of gown use in controlling vancomycin-resistant Enterococcus transmission: is it worth the price? Infect Control Hosp Epidemiol 2004; 25(5):418–424.

2 Outbreak Investigation in the Critical Care Unit Brian W. Cooper Division of Infectious Disease, Hartford Hospital, Hartford, and University of Connecticut School of Medicine, Farmington, Connecticut, U.S.A.

INTRODUCTION The term ‘‘outbreak’’ is a loaded word that tends to precipitate severe anxiety in hospital staff. Outbreaks of nosocomial infection, however, are quite common and often occur in critical care settings. An estimated 4% of all patients acquiring a nosocomial infection develop the infection as part of an outbreak (1). Outbreaks occur periodically in all institutions, and all clinicians should be aware of the potential dangers of outbreaks and the need for systematic surveillance and analysis of data in order to interdict clusters of infection as early as possible. Outbreaks are defined as an increase in the rate of occurrence of an event compared with past experience. The past experience is the baseline or endemic rate of event occurrence. Sometimes outbreaks are identified by large explosive increases in rates of infection. Such outbreaks are often easily identified. A three-fold increase in bacteremias due to Enterobacter cloacae during one month in an intensive care unit (ICU) would be a good example. Other outbreaks are more subtle and require careful analysis of surveillance data to recognize. For example, Kool et al. describe an outbreak of unrecognized hospital transmission of Legionnaires disease in transplant patients which ran over a two-year period before being recognized (2). Commonly, outbreaks may involve a single organism producing infections at a single site, as noted above. Other outbreaks involve clusters of particular types of infection caused by several different organisms. An example would be an increase in urinary tract infections arising due to poor technique in handling the catheters on a particular unit. Such an outbreak may involve infections due to several organisms. Sometimes outbreaks involve infections at several sites, as might happen with cutaneous infections, conjunctivitis, and bacteremia after introduction of Staphylococcus aureus into a neonatal ICU. Still other outbreaks may be quite difficult to detect. In particular, outbreaks of infections with long incubation periods are not easily discerned. Examples include nosocomial outbreaks of tuberculosis or hepatitis B viral infection. Outbreaks may also involve unusual sources of infection. For example,

33

34

Cooper

Gaillot et al. describe the spread of an extended spectrum beta lactamase (ESBL) producing Klebsiella strain, which was transmitted in ultrasonography gel (3).

SURVEILLANCE Surveillance of nosocomial infections provides the foundation for investigation of outbreak control efforts. It is the means by which a baseline or endemic rate is established. Surveillance of nosocomial infections involves systematic collection and analysis of data on the occurrence of infection. Care needs to be taken to make sure that well accepted definitions of nosocomial infection are followed by those conducting surveillance to ensure database validity. In addition, changes in surveillance methodology or casefinding methods will affect the comparability of data from one time period to another. When surveillance methods are significantly changed, data collected by the new surveillance activity can no longer be compared to data collected in prior time periods. Surveillance of nosocomial infections in hospitals is commonly carried out by infection control and hospital epidemiology personnel. Some critical care units may wish to supplement this surveillance activity with surveillance data of their own collected as part of a unit quality-improvement program. Such efforts need to be closely coordinated with infection-control personnel. Other sources of data in outbreaks include reports from clinicians or nursing units who may have noticed an unusual increase in particular infections. Mere collection of surveillance data is not sufficient in itself to detect clusters. The data must be analyzed periodically and compared with past experience. The most common means for this involves calculating a rate; one divides the crude number of infections by an appropriate denominator. Appropriate denominators are often derived from the patient census on a unit. The sensitivity of data analysis can be increased by using incidence density-adjusted denominators. For example, the rate of primary bacteremias in a given month or quarter may be expressed as a ratio of the number of infections to patients admitted to the unit. If, during a given quarter, six primary bacteremias are identified and 240 patients were admitted to the unit, the rate of infection may be expressed as 6/240 ¼ 3.3%. However, because some patients may be admitted to the unit only briefly while others have long-term stays, a more efficient denominator may be patient-days. When patient-days are used as a denominator, the rate is often multiplied by some constant such as 1000 to make the decimal point more manageable. In our example, if the 240 patients admitted to the unit accrued 1200 patient-days, the rate might be expressed as 6/1200  1000 ¼ 5 infections per 1000 patient-days. An even better way of calculating this rate would be to adjust for exposure to central intravenous catheters and use catheter-days as the denominator. Such rates more accurately reflect the exposure of patients to risk factors such as central intravenous catheters. Rates of infection collected in surveillance data should be compared with baseline rates using valid statistical methods to determine if a statistically significant increase in rate has occurred.

INVESTIGATION OF CLUSTERS OF INFECTION Outbreak investigation is a time- and resource-consuming process. Although there is no set script for all investigations, it is usually wise to proceed in a systematic fashion. Epidemiologists have derived a series of steps which comprise essential

Outbreak Investigation in the Critical Care Unit

35

Table 1 Systemic Steps in Outbreak Investigations Confirm the outbreak Establish a case definition and define the pre-epidemic and epidemic periods Notify appropriate hospital personnel Construct an epidemic curve to describe events in time, place, and person Review the literature Develop a line listing, chart review, and summary analysis of data Develop hypotheses and institute preliminary control measures Evaluate the effectiveness of control measures If necessary, proceed to further studies, such as case-control studies

milestones in the workup of an outbreak (Table 1). The order of these steps does not need to be followed precisely, but the first few steps listed are best completed initially. Confirming the Outbreak Because outbreak investigation is so resource-intensive, it is a waste of time to investigate small clusters that are not epidemics. When rates of infection are significantly elevated above baseline rates, an outbreak may be occurring. The investigator should next consider the possibility of a pseudo-outbreak. Pseudo-outbreaks usually occur when noninfecting organisms contaminate patient culture material. A good example has been described in the literature (4). Occasionally, pseudo-outbreaks occur when bias is introduced into case-finding methods, such as a change in definitions or surveillance methodology in the epidemic and pre-epidemic periods. Sometimes, changes in laboratory techniques during the pre-epidemic and epidemic periods may lead to pseudo-outbreaks. Establish a Case Definition In order to minimize bias in data collection, a precise written definition of cases needs to be developed. The definition should be broad enough that all potential cases are included yet not so broad as to bias the investigator with noncases. At times, the case definition may need to be redefined as the investigation proceeds. The definition should include the chief characteristics of the case diagnoses as well as appropriate factors indicating time, place, and person. An example of case definition for our fictitious outbreak of bacteremias may resemble the following: ‘‘A case was defined as any patient in the surgical ICU with Enterobacter cloacae isolated from blood cultures in the period after December 1, 1995.’’ All potential cases must fit the proposed definition or be rejected from the analysis. This allows uniformity in further data collection. The development of the case definition should also establish the pre-epidemic and epidemic periods. The epidemic period is usually the time from the first identified case to the present. The time period for the pre-epidemic period varies. In general, it is unwise to use a pre-epidemic period of less than six months. If the epidemic period is long or cases are few, then the pre-epidemic period should be lengthened. A one-year pre-epidemic period is commonly used in investigations of nosocomial outbreaks. NOTIFICATION OF APPROPRIATE INDIVIDUALS Good communication is essential in unraveling an outbreak investigation. Early in the investigation, the microbiology lab should be notified and the means developed to save outbreak organisms for possible epidemiologic typing later.

36

Cooper

Sometimes acute and convalescent serum needs to be collected and preserved. Numerous other details often need to be discussed with the laboratory director, and good lines of communication are essential. Affected chiefs of service, department heads, ICU directors, head nurses, and other clinical staff should be notified of the investigation. Often important information and direction can be gleaned by talking to these sources early on in the investigation. Outbreaks have political and sometimes public relations aspects as well. The hospital public relations director should be kept apprised of the investigation, but unsolicited press releases are usually not wise. When the press involves itself in an outbreak investigation and the investigators must talk to members of the media, it is best to be forthright and honest with information. All information released to the media should be channeled through a single hospital source to minimize confusion. Legal implications often flow from an outbreak investigation as well. At the outset, a new file should be opened and all decisions and meetings documented thoroughly. Outbreaks involving reportable diseases should, of course, be promptly reported to the state public health authorities. In some states, all outbreaks are required to be reported to the state. The list of reportable diseases also varies from state to state.

CONSTRUCT AN EPIDEMIC CURVE Epidemic curves are simple graphic tools which can convey surprising amounts of information concisely characterizing the outbreak. In an epidemic curve, individual cases are graphed over time. Additional data, such as mortality, location, or comorbid factors may be coded and superimposed on this curve. An additional graphic tool such as a spot map of a unit or wing may be useful in characterizing the epidemic’s geographic factors. Bed or room locations of cases can be easily displayed on the spot map.

REVIEW THE LITERATURE It is useful early in the investigation to gather information about the infection. The biology of the infecting organism, its reservoirs, and it mode of transmission should be carefully reviewed. Past outbreaks reported in the literature are important sources of information. Other investigations may have already dealt with similar situations, and solutions to your problems may have been suggested. Sources of information to search are myriad but should include computerized Medline searches, textbook sources, and Index Medicus searches. The medical librarian can be an invaluable resource to the investigator. We have found that a combination of computerized Medline searches and manual searches in textbooks and Index Medicus yields the best results.

DEVELOP A LINE LISTING A line listing is merely a questionnaire consisting of data from case records which the investigator deems potentially useful. Included are demographic data such as age, sex, and race, as well as patient diagnosis and location. Major diagnostic and

Outbreak Investigation in the Critical Care Unit

37

therapeutic procedures that the patient has undergone are also helpful. Underlying illnesses and medications administered are important to indicate. Listing the health care personnel taking care of the patient is often important, especially if a cluster of infections may be related to a health-care worker colonized with a pathogenic organism. Richman et al. (5) described an outbreak of Group A streptococcal surgical wound infections linked to exposure to an operating room staff member who was a rectal carrier of the organism. In addition, the line listing should consider all other potential risk factors for the illness noted by the investigator, including health-care or hygienic practices that may lead to illness in unusual ways. Claesson et al. (6) described an outbreak of Group A streptococcal endometritis on a maternity ward linked to use of a shower head. Data for the line listing are usually collected by chart review and stored in a computer database for ease of analysis. Preliminary analysis of the case is conducted by examining simple frequency rates and descriptions of the collected data. Clues to the outbreak may be apparent in the initial pattern of data collected. Risk factors that most cases have in common may be clues to solving an outbreak; however, in many cases the initial review of data is insufficient to establish the cause, and one must proceed to a case-control study as noted below.

DEVELOP HYPOTHESES AND INSTITUTE PRELIMINARY CONTROLS Even when initial data analysis is not sufficient to establish cause for the epidemic, the investigator commonly has developed early in the investigation several potential hypotheses regarding the underlying cause. Further analysis or data gathering may be necessary to prove out the likely hypothesis, but often at this stage some preliminary controls can be instituted. At times in an outbreak investigation, the preliminary control measures must be instituted early and empirically in the investigation due to the urgency of the situation. For example, in our institution, a case of nosocomial Salmonella gastroenteritis was noted in the newborn ICU (7). A quick review of the situation indicated that several staff members in the newborn ICU had been ill with mild gastrointestinal disturbance the week prior to the index case. While further investigation proceeded, preliminary controls were focused quickly on controlling exposure of newborns to potentially infected or colonized hospital staff. Ultimately, the investigation showed that no hospital staff were infected or colonized with S. gastroenteritis and that the infant was infected by her chronically colonized but asymptomatic mother. Control measures instituted early in the investigation should be based on hypotheses drawn from certain clues from the data gathered, which may suggest a source for the outbreak. For example, an outbreak of Serratia marcescens urinary tract infections among catheterized patients in an ICU might suggest problems in aseptic technique during catheter care. Control measures might include an education program stressing reinforcement of proper aseptic techniques in catheter care as well as redoubled efforts to encourage hand washing on the unit.

EVALUATE THE EFFECTIVENESS OF CONTROL MEASURES While control measures are being implemented, it is important to continue intensive surveillance for any new cases that continue to accrue. If new cases accumulate, one

38

Cooper

should question the effectiveness of the control measures. The investigator needs to be careful to take into account the incubation period of the infection during this period, however. Illnesses with relatively long incubation period such as a varicella may continue to accumulate after preliminary control measures were put into effect. FURTHER STUDIES If cases continue to accumulate beyond the institution of empiric controls, the investigator must question the accuracy of the hypothesis. Further hypotheses can be developed by evaluating multiple risk factors in a case-control study. In this type of study, case patients and appropriately chosen control patients are compared with regard to exposure to various risk factors. The investigator is looking for statistically significant association of certain risk factors with cases as opposed to controls. In an investigation of fever and hypotension on a surgical ICU, Trilla et al. (8) conducted a case-control study and found that volume of plasma expanders used was significantly associated with symptoms in case versus controls ( p ¼ .0029). Proper selection of control patients can be critical to the success of a casecontrol study. The art of selecting proper controls can be complicated, and the reader is referred to literature specifically dealing with this subject (9–11). Most importantly, care should be taken to ensure that cases and controls have an equal likelihood of being exposed to a set of risk factors. As a rule of thumb, three control patients should be chosen for each case patient. At the conclusion of the case-control study, a new understanding about relationships between risk factors and cases may suggest a need for a new set of control measures for the outbreak. SUMMARY Investigations of clusters or epidemics of nosocomial infection are often difficult, timeand resource-consuming activities. They tax the abilities and skills of infection-control personnel severely. However, outbreak investigations are among the most satisfying of infection control activities because of their far-reaching preventive effects. The proper response to a cluster of infections may mean the difference between a few cluster cases and a large-scale epidemic with significant morbidity and possible mortality. The reader is referred to the following references as examples of modern outbreak investigation. (12–18) REFERENCES 1. Wenzel RP, Thompson RL, Landry SM, et al. Hospital acquired infections in intensive care unit patients: an overview with emphasis on epidemics. Infect Control 1983; 4:371–375. 2. Kool JL, Fiore AE, Kioski CM, et al. More than 10 years of unrecognized nosocomial transmission of Legionnaires disease among transplant patients. Infect Control Hosp Epidemiol 1998; 19:905–910. 3. Gaillot O, Maruejouls C, Abachin E, et al. Nosocomial outbreak of Klebsiella pneumoniae producing SHV5 extended spectrum beta lactamase originating from a contaminated ultrasonography coupling gel. J Clin Microbiology 1998; 36:1357–1360. 4. Auerbach SB, McNeil MM, Brown JM, Lasker BA, Jarvis WR. Outbreak of pseudoinfections with Tsukamurella paurometabolum traced to laboratory contamination: efficacy of joint epidemiological and laboratory investigation. Clin Infect Dis 1992; 14:1015–1022.

Outbreak Investigation in the Critical Care Unit

39

5. Richman DD, Bretton SJ, Goldman DA. Scarlet fever and group A streptococcal surgical wound infection traced to an anal carrier. J Pediatr 1977; 90:387–390. 6. Claesson BE, Claesson UL. An outbreak of endometritis in a maternity ward caused by spread of group A streptococci from a shower head. J Hosp Infect 1985; 6:304–311. 7. Cooper, B. Unpublished observation. 8. Trilla A, Codina C, Salles M, et al. A cluster of fever and hypotension on a surgical intensive care unit related to the contamination of plasma expanders by cell wall products of Bacillus stearothermophilus. Infect Control Hosp Epidemiol 1995; 16:335–339. 9. Cole P. The evolving case control study. J Chron Dis 1979; 32:15–27. 10. Feinstein AR. Experimental requirements and scientific principles in case control studies. J Chron Dis 1985; 38:127–133. 11. Hayden GF, Kramer MS, Horwitz RI. The case control study: a practical review for the clinician. JAMA 1982; 247:326–331. 12. Pimental JD, Low J, Styles K, et al. Control of an outbreak of multi-drug-resistance Acinetobacter baumannii in an intensive care unit and a surgical ward. J Hosp Infect 2005; 59(3):249–253. 13. Jeong SH, Bae IK, Kwon SB, et al. Investigation of a nosocomial outbreak of Acinetobacter baumannii producing PER-1 extended spectrum beta lactamase in an intensive care unit. J Hosp Infect 2005; 59(3):242–248. 14. Qavi A, Segal-Maurer S, Mariano N, et al. Increased mortality associated with a clonal outbreak of ceftazidime-resistant Klebsiella pneumoniae: a case-control study. Infect Control Hosp Epidemiol 2005; 26(1):63–68. 15. Zawacki A, O’Rourke E, Potter-Bynoe G, et al. An outbreak of Pseudomonas aeruginosa pneumonia and bloodstream infection associated with intermittent otitis externa in a healthcare worker. Infect Control Hosp Epidemiol 2004; 25(12):1083–1089. 16. Behari P, Englund J, Alcasid G, et al. Transmission of methicillin-resistant Staphylococcus aureus to preterm infants through breast milk. Infect Control Hosp Epidemiol 2004; 25(9):778–780. 17. Bukholm G, Tannaes T, Kjelsberg AB, et al. An outbreak of multidrug-resistant Pseudomonas aeruginosa associated with increased risk of patient death in an intensive care unit. Infect Control Hosp Epidemiol 2002; 23(8):441–446. 18. Lehours P, Rogues AM, Occhialini A, et al. Investigation of an outbreak due to Alcaligenes xylosoxidans subspecies xylosoxydans by random amplified polymorphic DNA analysis. Eur J Clin Microbiol Infect Dis 2002; 21(2):108–113.

3 Clinical Approach to Fever in the Critical Care Unit Burke A. Cunha Infectious Disease Division, Winthrop-University Hospital, Mineola, and State University of New York School of Medicine, Stony Brook, New York, U.S.A.

INTRODUCTION Overview Fever is a cardinal sign of disease. Fever may be caused by a wide variety of infectious and noninfectious disorders. The number of disorders that occur in seriously ill patients in critical care unit (CCU) is more limited than in the non-CCU population. The main clinical problems in the CCU are to differentiate between noninfectious and infectious causes of fever rather than to try and determine the specific cause of the patient’s fever. The clinical approach to fever in the CCU is based on a careful analysis of the acuteness/chronicity of the fever, the characteristics of the fever pattern, the relationship of the pulse to the fever, the duration of the fever, and the defervescence pattern of the fever. It is the task of the infectious disease consultant to relate aspects of the patient’s history and physical, laboratory, and radiological tests with the characteristics of the patient’s fever, which together determine differential diagnostic possibilities. After the differential diagnosis has been narrowed by analysis of the fever’s characteristics and the patient-related factors mentioned, it is usually relatively straightforward to order tests to arrive at a specific diagnosis. Most patients in the CCU have some degree of temperature elevation. Trying to determine the cause of fever in CCU patients is the daily task of the physicians concerned. Fever in the CCU can be a perplexing problem, because the clinician must determine whether the patient’s underlying disorder is responsible for the fever or fever is a superimposed phenomenon on the patient’s underlying problem responsible for admission to the CCU. The infectious disease consultant’s clinical skills are best demonstrated by the rapidity and excellence of arriving at a cause for the patient’s fever (1–10).

41

42

Cunha

DIAGNOSTIC CONSIDERATIONS The clinician’s initial assessment of the febrile patient is based on the temperature elevation/patterns, underlying disorders, assessment of medications, and clinical appearance. CCU patients may be grouped into three categories. The first group of patients is the one that has been admitted to the hospital and directly put in the CCU. Such patients need immediate/appropriate diagnostic tests, and if an infectious etiology is likely, empiric antibiotic therapy based on organ system involvement is needed. Some patients with lymphoma, vasculitis, or systemic lupus erythematosus (SLE) may be febrile/critically ill and in the CCU, but these disorders usually have a less fulminant clinical presentation. The second group of CCU patients is those transferred to the CCU after hospitalization. This group of patients has been hospitalized for a variety of reasons, and some catastrophic event during hospitalization requires transfer to the CCU for intensive evaluation/organ support. The usual infectious diseases in this group include nosocomial pneumonia, intravenous (IV)-line infections, Clostridium difficile colitis, postoperative rupture of a viscus, or leakage of a partially drained/undrained abscess. Equally important are noninfectious diseases occurring in hospitalized patients who require transfer to the CCU. Commonly, these include acute myocardial infarction, pulmonary emboli/infarcts, acute pancreatitis, acute adrenal insufficiency and internal/gastrointestinal hemorrhage. Unless urologically instrumented during hospitalization, urosepsis is rare; but if there is renal disease, SLE, or diabetes, it may occur. The third group of patients has underlying infectious/noninfectious disorders that may flare and be superimposed on the patient’s underlying medical disorders. Most disorders in this group are noninfectious diseases and present in the stabilized hospital patient with a new fever. Such disorders include an acute attack of gout, thrombophlebitis, phlebitis, C. difficile diarrhea, or connective tissue diseases (usually SLE or rheumatoid arthritis), atelectasis/dehydration, pleural effusions, and drug fever (1,5,6,10).

CAUSES OF FEVER IN THE CCU Noninfectious Causes of Fever in the CCU A wide variety of disorders are associated with a febrile response. Both infectious and noninfectious disorders may cause acute/chronic fevers that may be low, i.e., 102 F, or high grade, i.e., 102 F. Of the multiplicity of conditions that may be encountered in the CCU with a few notable exceptions, most noninfectious disorders are associated with fevers of 102 F. Exceptions to the 102 F fever rule include malignant hyperthermia, adrenal insufficiency, massive intracranial hemorrhage, central fever, drug fever, collagen vascular disease flare, particularly SLE flare, heat stroke, vasculitis, and certain malignancies, particularly lymphomas. The most common noninfectious disorders encountered in the CCU either have no fever or have low-grade fevers 102 F, and include acute myocardial infarction, pulmonary embolism/infarct, phlebitis, catheter-associated bacteriuria, acute pancreatitis, viral hepatitis, acute hepatic necrosis, dry gangrene, uncomplicated wound infections, subacute bacterial endocarditis, cerebrovascular accidents, smallmoderate intracerebral bleeds, pulmonary hemorrhage, acute respiratory distress syndrome (ARDS), bronchiolitis obliterans with organizing pneumonia (BOOP),

Clinical Approach to Fever in the Critical Care Unit

43

pleural effusions, atelectasis, cholecystitis, noninfectious diarrheas, C. difficile diarrhea, ischemic colitis, splenic infarcts, renal infarcts, pericardial effusion, dry gangrene, gas gangrene, surgical toxic shock syndrome, acute gout, small bowel obstruction, and cellulitis. The clinical approach to the noninfectious disorders with fever is usually relatively straightforward, because they are readily diagnosable by history, physical, or routine laboratory or radiology tests. Having known that noninfectious disorders are not associated with fevers >102 F in patients, the clinician can look for an alternate explanation in these patients. The difficulty usually arises when the patient has a multiplicity of conditions and sorting out the infectious from the noninfectious causes can be a daunting task. For example, if a patient in the CCU following cancer resection of the large bowel postoperatively develops extremity gangrene, phlebitis, pulmonary infiltrates, leukocytosis, myocardial infarction, hematomas/ seromas, atelectasis, dehydration, and catheter-associated bacteriuria, has a stroke, is on multiple medications, and has just had a blood transfusion, the clinical analysis of the fever in this patient would go as follows. The patient spiked to 102.8 F on the seventh postoperative day and 24 hours after receiving the blood transfusion. Postoperatively, if there was no peritonitis, the temperatures due to hematoma/healing should be 102 F, they have the potential to do so, e.g., nosocomial pneumonia may be associated with temperatures 102 F. Although all infectious diseases will not present with temperatures 102 F, they are the disorders most frequently associated with temperatures in the 102 F range. Infectious diseases encountered in the CCU usually associated with temperatures 102 F include postoperative abscesses, acute meningitis, acute encephalitis, brain abscess, septic thrombophlebitis, jugular septic vein thrombophlebitis, septic pelvic thrombophlebitis, septic pulmonary emboli, pericarditis, acute bacterial endocarditis, perivalvular/myocardial abscess, community-acquired pneumonia (CAP), pleural empyema, lung abscess, cholangitis, intrarenal/perinephric abscess, prostatic abscess, urosepsis, central-line infections, contaminated infusates, pylephlebitis, liver abscess, C. difficile colitis, complicated skin and soft tissue infections/abscesses, AV graft infections, foreign body–related infections (infected pacemakers/automatic implantable cardioveter defibrillator (AICD)s, central IV catheters, Hickman/ Broviac catheters), and septic arthritis. Infectious diseases likely to be present in the CCU setting with temperatures 106 F) include heat stroke, malignant hyperthermia, central fevers, malignant neoplastic syndrome, and drug fever. These hyperpyrexia disorders are usually readily diagnosed because of associated features, i.e., fever immediately following surgery/general anesthesia (malignant hyperthermia), prolonged high temperatures/dehydration (prolonged heat exposure), central nervous system (CNS) disorders with hypothalamic involvement (tumors, neurosurgical procedures, and trauma), or sensitivity medications (drug fever). Temperatures in excess of 106 F should suggest a noninfectious etiology (Table 3) (1,10,11). Single Fever Spikes >102 F Patients in the CCU who have been afebrile or had low-grade fevers, i.e., 102 F, may suddenly develop a single fever spike >102 F. Single fever spikes are never infectious in origin. The causes of single fever spikes include insertion/removal of a urinary catheter or a venous catheter, suctioning/manipulation of an endotracheal tube, wound packing/lavage, wound irrigation, etc. Any procedure that involves a manipulation of a colonized/infected surface can induce a transient bacteremia. Because of their short duration, i.e., less than five minutes, such bacteremias do not result in sustaining infection or spread infection to other organs, and for this reason may not be treated. Single fever spikes of the transient bacteremias are a diagnostic not a therapeutic problem. The other common cause of single fever spikes in the CCU is blood-product transfusions. Fever secondary to blood products/blood transfusions are a frequent occurrence, and are most commonly manifested by fever following the infusion. The distribution of fever is bimodal, following a blood transfusion. Most reactions occur within the first 72 hours after the blood/blood product transfusion, and most reactions within the 72-hour period occur in the first 24 to 48 hours. There are very few reactions after 72 hours, but a smaller peak five to seven days after the blood transfusion, which although is very uncommon, may occur. Temperature elevations associated with late blood transfusion reactions are lower than those with reactions occurring soon after blood transfusion. The fever subsequent to the transient bacteremia results from cytokine release and is not indicative of a prolonged exposure to the infecting agent, but rather represents the postbacteremia chemokine-induced febrile response. The temperature elevations from manipulation of a colonized or infected mucosal surface persist long after the bacteremia has ceased.

46

Cunha

Table 2 Clinical Applications of the 102 F Rule in the Critical Care Unit Common causes of temperature Acute myocardial infarction Pulmonary embolism

GI bleed

Acute pancreatitis Hematomas Phlebitis Catheter-associated bacteriuria

Pleural effusions

Uncomplicated wound infections Atelectasis/dehydration Tracheobronchitis

Thrombophlebitis

Clostridium difficile

Nosocomial pneumonia/ ventilator-associated pneumonia

Comments 102 F H/O chest pain/CAP EKG/cardiac enzymes H/O PE underlying reasons predisposing to pulmonary emboli VQ scan positive (pulmonary angiography for large emboli) " FSPs with multiple small pulmonary emboli Hyperactive BS, BRB per rectum/melena " BUN (except in alcoholic liver disease) Endoscopy/abdominal CT scan ! bleeding source Severe abdominal pain: may be associated with ARDS " amylase and " lipase or pancreatitis on abdominal CT scan H/O recent surgery/bleeding diathesis Visible on skin, e.g., Grey-Turner’s/Cullen’s sign, or on CT scan Local erythema without suppuration/vein tenderness Urine with bacteria and WBCs nearly always represent colonization, not infection Bacteremia (urosepsis) does not result from bacteriuria unless pre-existing renal disease, urinary tract obstruction, or patient has SLE, DM, steroids, etc. Bilateral effusions are never due to infection: look for a noninfectious etiology Except for gas gangrene and streptococcal cellulitis, temperatures are usually low grade ‘‘Wounds’’ with temperatures 102 F should prompt a search for an underlying abscess Temperatures usually 101 F May be confused with pulmonary emboli/early pneumonia Purulent endotracheal secretions with negative chest X ray for pneumonia Tracheobronchitis does not ! temperatures 102 F Warm, tender calf/foot veins  palpable cord Thrombophlebitis does not ! pulmonary emboli Phlebothrombosis ! pulmonary emboli Stools positive for C. difficile toxin Fecal WBC positive 50%. Stools are watery, green, foul smelling If temperatures 102 F with blood/mucous diarrhea, diagnosis is probable >102 F May have temperatures 102 F also

Pulmonary infiltrate consistent with a bacterial pneumonia occurring 1 wk after hospitalization Must be differentiated from ARDS, LVF, etc. Definitive diagnosis by quantitative protected catheter-tip culture (PBB/BAL) Endotracheal secretions represent upper airway colonization and are not reflective of lower respiratory tract organisms causing vent-associated pneumonia (Continued )

Clinical Approach to Fever in the Critical Care Unit Table 2

47

Clinical Applications of the 102 F Rule in the Critical Care Unit (Continued )

Common causes of temperature

IV-line infections

C. difficile

Drug fever

Blood/blood product transfusion Transient bacteremia due to manipulation of a colonized/infected mucosal surface September invasive infectious diseases

Comments Endotracheal respiratory secretion cultures should not be cultured or covered Overdue central lines usual cause Organisms from blood cultures taken from noninvolved extremity same as positive semiquantitative catheter culture (15 colonies) If all other sources of fever are ruled out, consider IV-line infection, especially with overdue lines (even if site not infected visually) Stools positive for C. difficile toxin Bloody diarrhea, temperature 102 F Abdominal CT scan shows thumbprinting/colitis/toxic megacolon In patients with otherwise unexplained temperatures, consider drug fever Blood cultures are negative (excluding contaminants) Patients with drug fever usually have 102 F accompanied by relative bradycardia " WBC with left shift is common as is " ESR Mild-moderate serum transaminases common Eosinophils nearly always present but eosinophils less common Patient with infection may also have a drug fever Commonest causes of drug fever are diuretics, pain/sleep medications, sulfa-containing stool softeners as well as sulfa drugs, and b-lactam antibiotics Single fever spike (1–3 or 5–7 days posttransfusion) Single temperature spike 1–3 days, postmanipulative, that spontaneously resolves without treatment

Blood culture variably positive as a function of time Usually associated with temperatures 102 F in normal hosts

Abbreviations: GI, gastrointestinal; CT, computed tomography; ARDS, acute respiratory distress syndrome; SLE, systemic lupus erythematosus; IV, intravenous; WBC, white blood cells; DM, diabetes mellitus; PBB, protected bronchial brushings; BAL, bronchoalveolar lavage; ESR, erythrocyte sedimentation rate; CAP, community-acquired pneumonia; PE, pulmonary emboli; FSPs, fibrin split products. Source: Adapted from Refs. 3, 5.

Table 3 Causes of Extreme Hyperpyrexia (>106 F) Hypothalamic dysfunction Malignant neuroleptic syndrome Central fevers (hemorrhagic, trauma, malignancy) Heat stroke Malignant hyperthermia Drug fever

48

Cunha

In patients with fever spikes due to transient bacteremias following manipulation of a colonized or infected mucosal surface, or secondary to a blood product/ blood transfusion, may be inferred by the temporal relationship of the event and the appearance of the fever. In addition to the temporal relationship between the fever and the transient bacteremia or transfusion-related febrile response, the characteristic fever curve, i.e., a single isolated temperature spike resolves spontaneously without treatment (1,6,10,11). Multiple Fever Spikes >102 F Multiple fever spikes >102 F may be infectious or noninfectious in origin, because a hectic, septic fever pattern does not in itself suggest a particular etiology. Because this fever pattern is common but not specific, the clinician must rely upon associated findings in the history and physical, or among laboratory or radiology tests, to determine the cause of the fever. Pulse temperature relationships are also of help in differentiating the causes of fever in patients with multiple temperature spikes over a period of days. Assuming that there is no characteristic fever pattern, the presence or absence of a pulse temperature deficit is useful. Patients with a pulse temperature deficit, i.e., relative bradycardia, are limited to relatively few infectious and noninfectious disorders. In the CCU setting, patients with multiple spiking fevers and a pulse temperature deficit should suggest Rocky Mountain spotted fever (RMSF), typhus, arboviral hemorrhagic fevers, central fevers, lymphoma-related fevers, legionnaires’ disease, Q fever, psittacosis, or drug fever. The diagnostic significance of relative bradycardia can only be applied in patients who have normal pulse temperature relationships, i.e., those who do not have pacemaker-induced rhythms, have third-degree heart block, those with arrhythmias, and those on b-blocker therapy. Any patient on b-blocker medications who develops a fever will develop relative bradycardia, thus eliminating the usefulness of this important diagnostic sign in patients with relative bradycardia. If these conditions are eliminated, narrowing diagnostic possibilities is relatively straightforward. If the patient has pneumonia and relative bradycardia, then Legionella, psittacosis, or Q fever is suggested. Patients with relative bradycardia may accompany lymphoma or central fever, and in those conditions should not suggest an alternative diagnosis. Typhus or RMSF will be suggested by the pattern/nature of their rash as well as other findings, and relative bradycardia is an ancillary finding in those situations. Similarly, the arboviral hemorrhagic fevers do not require the presence of a pulse temperature deficit for diagnosis. In those situations, epidemiology/recent travel history, systemic toxemia, and hemorrhagic nature of the rash are the primary diagnostic determinants (1,5,10,11). Drug fever, i.e., a hypersensitivity reaction to medications without rash, occurs in approximately 10% of hospitalized patients. In the CCU where multiple medications are being used, drug fever is a common cause of obscure fever. Drug fever is a diagnosis of exclusion, and its presence is suggested by the presence of relative bradycardia in the absence of another explanation for the fever. Drug fever with relative bradycardia may also be a problem that is related to the therapy of the underlying disorder that prompted CCU admission. Aside from being a diagnosis of elimination, drug fever is suggested by negative blood cultures excluding contaminants, increase in the erythrocyte sedimentation rate (ESR), and an early mild/transient increase in the serum transaminases serum glutamate oxaloacetate transaminase/ serum glutamate pyruvate transaminase (SGOT/SGPT). Drug fevers 102 F for more than three days should suggest a complication or alternate diagnosis. Other disorders with prolonged low-grade fevers include dehydration, atelectasis, wound healing, hematoma, seromas, ARDS, BOOP, deep vein thromboses, pleural effusions, tracheobronchitis, decubitus ulcers, cellulites, and phlebitis. Prolonged low-grade fevers (102 F) are usually noninfectious. Clinicians should try to identify the noninfectious disorder causing the fever so that undue resources will not be expended looking for an infectious explanation for the fever (1,10,17–24). Low-Grade Prolonged Fevers (more than five days) in the CCU: Nosocomial FUOs There are relatively few causes of prolonged fevers in the CCU that last for over a week. Such low-grade prolonged fevers lasting over a week have been termed nosocomial fevers of unknown origin (FUO). There are relatively few causes of nosocomial FUO in contrast to its community-acquired counterpart. Low-grade infections or inflammatory states account for most of the causes of nosocomial FUO. Nosocomial FUOs are usually due to central fevers, drug fevers, postperfusion syndrome, atelectasis, dehydration, undrained seromas, tracheobronchitis, and catheter-associated bacteriuria. Prolonged fevers that become high spiking fevers should suggest the possibility of nosocomial endocarditis related to a central line or invasive cardiac procedure. Prolonged high spiking fevers can also be due to septic thrombophlebitis or an undrained abscess. Nosocomial sinusitis due to prolonged nasotracheal intubation is a rare cause of nosocomial FUO (1,10,25). Special Fever Patterns Camel Back/Domedary Fevers Fever patterns are often considered as nonspecific in nature therefore having limited diagnostic specificity. It is true that fevers in patients being intermittently given antipyretics and being instrumented in a variety of anatomical locations, do have complex fever patterns. However, these are usually easily sorted out on the basis of clinical findings. Fever patterns that remain useful to diagnose certain entities in the hospital include a ‘‘domedary’’ or ‘‘camel back’’ fever pattern, i.e., increase in fever over two to three days followed by a decrease followed a few days later by two to three more

50

Cunha

days of fever. Such a fever pattern should suggest the possibility of Colorado tick fever, dengue, leptospirosis, brucellosis, lymphocytic choriomeningitis, yellow fever, the African hemorrhagic fevers, rat bite fever, and smallpox. Sustained/Remittent Fevers Continuous fevers are those that rise slowly over days to peak levels and then plateau until defervescence. The causes of continuous/sustained fevers include typhoid fever, drug fever, scarlet fever, RMSF, psittacosis, Kawasaki’s disease, brucellosis, human herpes virus six (HHV-6) infections, and central fevers. Remittent fevers are characteristic of viral respiratory tract infection, malaria, acute rheumatic fever, legionnaires’ disease, Legionella/mycoplasma CAP, tuberculosis (TB), and viridans streptococcal subacute bacterial endocarditis (SBE). Hectic/Septic Fevers Hectic/septic fevers are repetitive fever spikes over days that may or may not decrease to normal between fever spikes. Hectic/septic fevers may be due to gramnegative or gram-positive sepsis, renal, abdominal, or pelvic abscesses, acute bacterial endocarditis, Kawasaki’s disease, malaria, miliary TB, peritonitis, or toxic shock syndrome. Noninfectious causes include lymphomas or overzealous administration of antipyretics. Double Quotidian Fevers Double quotidian fevers, i.e., two fever spikes in 24 hours, not artificially induced by antipyretics, should suggest right-sided gonococcal endocarditis, mixed malarial infections, miliary TB, visceral leishmaniasis, or adult Still’s disease. Double quotidian fevers are helpful, when present, in narrowing diagnostic possibilities enabling the clinician to order specific diagnostic testing for likely diagnostic possibilities (Table 4) (1,3,10,11). Relapsing Fevers Relapsing fevers are those that are separated by afebrile intervals. Causes of noninfectious relapsing fevers include Crohn’s disease, Behc¸et’s disease, relapsing funiculitis, leukoclastic angiitis, Sweet’s syndrome, familial Mediterranean fever, fever aphthous ulcer pharyngitis adenopathy (FAPA) syndrome, hyper IgG syndrome, and SLE. The infectious causes of relapsing fevers include viral infections, i.e., cytomegalovirus (CMV), Epstein–Barr virus, lymphatic choriomeningiti virus (LCM), dengue, yellow fever, and Colorado tick fever. Zoonotic bacterial infections include leptospirosis, bartonellosis, brucellosis, rat bite fever (S. minus), visceral leishmaniasis, malaria, babesiosis, ehrlichiosis, Q fever, typhoid fever, trench fever, and relapsing fever. Fungal infections tend to relapse, as do melioidosis and TB. Chronic meningococcemia by definition is prone to relapse (Table 5) (1,3,10,11). PULSE–TEMPERATURE RELATIONSHIPS Diagnostic Importance of Relative Bradycardia Relative bradycardia is a pulse temperature deficit inappropriate for the degree of fever. A temperature increase of normally 1 F is accompanied by an appropriate

Clinical Approach to Fever in the Critical Care Unit

51

Table 4 Diagnostic Significance of Fever Patterns Fever pattern Single fever spike

Intermittent (hectic/septic) fevers

Remittent fevers

Continuous/sustained fevers

Double quotidian fevers

Biphasic (Camelback) fevers

Usual causes Manipulation of a colonized/infected mucosal surface (not systemic infectious disease) Blood/blood products transfusion Infusion-related sepsis (contaminated infusate) Temperature error Gram-negative/positive sepsis Abscesses (renal, abdominal, pelvic) Acute bacterial endocarditis Kawasaki disease Malaria Miliary TB Peritonitis Toxic shock syndrome Antipyretics Viral upper respiratory tract infections Plasmodium falciparum malaria ARF Legionella Mycoplasma TB SBE (viridans streptococci) Central fevers Roseola infantum (human herpesvirus-6) Brucellosis Kawasaki disease Psittacosis RMSF Scarlet fever Enterococcal SBE (tularemia) Typhoid fever Drug fever Adult Still’s disease (adult–juvenile rheumatoid arthritis) Visceral leishmaniasis Miliary TB Mixed malarial infections Right-sided gonococcal endocarditis Colorado tick fever Dengue fever Leptospirosis Brucellosis Lymphocytic choriomeningitis Yellow fever Poliomyelitis Smallpox Rat-bite fever (Spirillum minus) Chikungunya fever Rift Valley fever (Continued )

52 Table 4

Cunha Fevers Prone to Relapse (Continued )

Fever pattern

Relapsing fevers

Usual causes African hemorrhagic fevers (Marburg, Ebola, Lassa, etc.) Echovirus (Echo 9) Relapsing fever (B. recurrentis) Yellow fever Smallpox Ascending (intermittent) cholangitis Brucellosis Dengue Chronic meningococcemia Malaria Rat-bite fever (S. moniliformis)

Abbreviations: RMSF, Rocky Mountain spotted fever; SBE, subacute bacterial endocarditis; TB, tuberculosis; ARF, acute rheumatic fever. Source: From Ref. 11.

pulse increase of 10 beats/min. If the pulse response is less than it should be for any degree of temperature increase ( >102 F), then the term relative bradycardia may be applied. Relative bradycardia combined in a patient with an obscure fever is an extremely useful diagnostic sign. Fever plus relative bradycardia immediately limits

Table 5 Fevers Prone to Relapse Infectious causes Relapsing fever (Borrelia recurrentis) Trench fever (Rochalimaea quintana) Q fever Typhoid fever Vibrio fetus Syphilus TB Histoplasmosis Coccidioidomycosis Blastomycosis Pseudomonas pseudomallei (meliodosis) LCM Dengue fever Yellow fever Chronic meningococcemia Noninfectious causes Behcet’s disease Crohn’s disease Weber-Christian disease (panniculitis) Leukoclastic angitis Sweet’s syndrome

Colorado tick fever Dengue fever Leptospirosis Brucellosis Bartonellosis (Oroyo fever) Acute rheumatic fever Rate-bite fever (Spirillum minus) Visceral leishmaniasis Lyme disease Malaria Noninfluenzal respiratory viruses Babesiosis EBV CMV

Familial Mediterranean fever Fever, adenitis, pharyngitis, aphthous ulcer syndrome SLE Hyper IgD syndrome

Abbreviations: EBV, Epstein–Barr virus, CMV, cytomegalovirus; SLE, systemic lupus erythematosus; TB, tuberculosis; LCM, lymphatic choriomeningitis virus. Source: From Ref. 11.

Clinical Approach to Fever in the Critical Care Unit

53

diagnostic possibilities to central fevers, drug fevers, lymphomas, and the noninfectious disorders commonly causing fever in the CCU. Among the infectious causes of fever in the CCU, relative bradycardia in patients with pneumonia narrows diagnostic possibilities to Legionella, psittacosis, or Q fever pneumonia. Patients without pneumonias with fevers in the CCU, limit diagnostic possibilities to a variety of arthropod-borne infections, i.e., RMSF, typhus, typhoid fever, and arthropod-borne hemorrhagic fevers, i.e., yellow fever, Ebola, and dengue fever. Relative bradycardia like other signs should be used in concert with other clinical findings to prompt further diagnostic testing for the infectious diseases and to eliminate from further consideration the noninfectious disorders associated with relative bradycardia (Table 6; Fig. 1) (1,10,26,28–31).

Table 6 Causes of Relative Bradycardia Infectious Legionella Psittacosis Q fever Typhoid fever Typhus Babesiosis Malaria Leptospirosis Yellow fever Dengue fever Viral hemorrhagic fevers RMSF

Noninfectious Beta blockers Verapamil, or diltiazem CNS lesions Lymphomas Factitious fever Drug fever

Determination of relative bradycardia Inclusive criteria Patient must be an adult, i.e., >13 yr Temperature >102 F Pulse must be taken simultaneously with the temperature elevation Exclusive criteria Patient has NSR without arrhythmia, second-or third-degree heart block or pacemaker-induced rhythm Patient must not be on b-blocker medications Appropriate Temperature-Pulse Relationships Pulse (beats/min) Temperature 150 41.1 C (106 F) 140 40.6 C (105 F) 130 40.7 C (104 F) 120 39.4 C (103 F) 110 38.9 C (102 F) Relative bradycardia refers to heart rates that are inappropriately slow relative to body temperature (pulse must be taken simultaneously with temperature elevation). Applies to adult patients with temperature 102 F; does not apply to patients with second/third-degree heart block, pacemaker-induced rhythms, or those taking beta-blockers. Abbreviations: CNS, central neurous system; RMSF, Rocky Mountain spotted fever; NSR, normal sinus rhythm.

54

Cunha

SHAKING CHILLS HEADACHE DIARRHEA 1

DAY

MENTAL CONFUSION 2

4

3

5

6

7

8

9

10

12

11

13

14

15

PULSE TEMP. 130

104

120

103

110

102

100

101

90

100

80

99

78

98

60 97 CEFAMANDOLE 2gm (IV) 96h ANCEF 1gm (IV) 500gm (IV) AMIKACIN DOXYCYCLINE LABS 16.0 13.1 WBC 2.5 PO4 1.3 1.2 CREAT.

200mg (IV) g12h 6.8 10.8

10.9 2.4 1.2

1.2 1.0

100 mg (IV) g12h 100 mg (PO) g12h 9.0

9.2

10.4

0.7

Figure 1 Temperature chart showing relative bradycardia in a patient with legionnaires’ disease prior to initiation of doxycycline treatment on day 5. Solid line represents temperature; dotted line represents pulse. Source: From Ref. 27.

DIAGNOSTIC SIGNIFICANCE OF FEVER DEFERVESCENCE PATTERNS Overview Most of this chapter has been concerned with the diagnosis of fever in the CCU by analyzing the rapidity of onset of the fever, the height of the fever, the relationship of the fever to the pulse, the fever patterns, and the duration of the fever. Particularly in perplexing cases of fever, the characteristics of fever resolution also have diagnostic significance. Fever defervescence patterns may be interpreted in two ways. The rapidity and completeness of the fever pattern resolution attests to the effective treatment or resolution of the noninfectious or infectious process. Fever defervescence patterns are as predictable and useful as fever patterns in predicting complications secondary to the disorder or therapy (5,10,11).

Meningitis in the CCU With bacterial meningitis, temperature resolution with appropriate therapy is related to the pathogen causing the meningitis. Meningococcal meningitis defervesces quickly over one to three days, whereas Haemophilus influenzae meningitis resolves over three to five days, and severe pneumococcal meningitis may take a week or longer for the fever to decrease/become afebrile. Viral causes of meningitis or encephalitis defervesce very slowly over a seven-day period, and by fever defervescence pattern easily differentiate viral meningitis/encephalitis from bacterial meningitis. Because fever defervescence patterns may also point to complications, the astute clinician will monitor the fever pattern post-therapy, looking for an unexpected temperature spike after the patient has defervesced.

Clinical Approach to Fever in the Critical Care Unit

55

H. influenzae meningitis, for example, defervesces after three to five days but if the patient spikes a temperature after five days, this would suggest either a complication of the infection, i.e., subdural empyema, or a complication of therapy, i.e., drug fever secondary to antimicrobial therapy (1,5,10). Endocarditis in the CCU In patients with endocarditis, the fever defervescence pattern is also pathogen related. Patients with SBE have fevers 102 F, but fever is not helpful in ruling in or out the diagnosis of nosocomial pneumonia. The NP is an imprecise diagnosis and is routinely given to most patients in the CCU who have fever,

56

Cunha

leukocytosis, and pulmonary infiltrates. Therefore, most patients who have a working diagnosis of NP in fact do not have NP but have infiltrates, fever, and leukocytosis due to other causes. Patients being treated appropriately with monotherapy or combination therapy for NP defervesce rapidly if the infiltrates do in fact represent NP. Monotherapy or combination therapy for NP should be with at least one agent that has a high degree of anti–P. aeruginosa activity. Patients with bona fide NP defervesce within a week (32–37). The persistence of fever, i.e., lack of a fever in patients with possible NP, suggests two possibilities; firstly, the patient has a noninfectious disorder that is mimicking NP and for this reason is not responding to antimicrobial therapy. Secondly, the patient could have an infectious disease, a process that is unresponsive to antipseudomonal antimicrobial therapy, i.e., herpes simplex virus 1 (HSV-1) pneumonia. HSV-1 pneumonia is common in the CCU setting and presents as persistent fever and infiltrates unresponsive to antibiotics, or as ‘‘failure to wean’’ in ventilated patients. Patients who present as ‘‘failure to wean’’ have persistent fevers and did not have antecedent severe lung disease that would compromise their ability to come off the respirator. NP with empiric treatment should see an improvement/resolution of infiltrates and a defervescence of fever within two weeks. Persistence of fever with or without infiltrates after two weeks, in the absence of another cause for the fever, should suggest HSV-1 pneumonia until proven otherwise. HSV-1 pneumonia is diagnosed by bronchoscopy, demonstrating cytopathic effects from cytology specimens, or direct fluorescent antibody (DFA)/monoclonal tests of respiratory secretions will be positive for HSV. Importantly, no vesicles are present in the bronchi in bronchoscoped patients with HSV-1 pneumonitis (38,39).

OBSCURE FEVERS IN THE CCUs Drug Fever Drug fevers are so important in the CCU setting because of the multiplicity of medications. Physicians should always be suspicious of the possibility of drug fever when other diagnostic possibilities have been exhausted. Drug fever may occur in individuals who have just recently been started on the sensitizing medication, or more commonly who have been on a sensitizing medication for a long period of time without previous problems. Patients with drug fever do not necessarily have multiple allergies to medications, and are not usually atopic. However, the likelihood of drug fever is enhanced in patients who are atopic with multiple drug allergies. Patients with drug fever, i.e., hypersensitivity reaction without rash may present with any degree of fever, but most commonly drug fevers are in the 102 F to 104 F range. Other conditions aside, patients look ‘‘inappropriately well’’ for the degree of fever that is different than the toxemic patient with a serious bacterial systemic infection. Relative bradycardia is invariably present excluding patients on b-blocker therapy, those with arrhythmias, heart block, or pacemaker-induced rhythms. Laboratory tests include an increase in WBC count with a shift to the left. Eosinophils are often present early in the differential count, but less commonly is their actual eosinophilia. The ESR is increased with drug fever but this may be marked by other causes of increased ESR by one/more acute disorders in CCU patients. The sedimentation rate also is increased after surgical procedures, negating the usefulness of this test in the postoperative fever patient. Serum transaminases, i.e., SGOT/SGPT are also mildly/transiently elevated early in cases of drug fever. Such mild increases in the

Clinical Approach to Fever in the Critical Care Unit

57

serum transaminases are often overlooked by clinicians as acute phase reactants or not being very elevated. However, in a patient with an obscure otherwise unexplained fever the constellation of nonspecific findings including relative bradycardia, slightly increased serum transaminases, and eosinophils in the differential count, are sufficient to make a presumptive diagnosis of drug fever. It is a popular misconception that antibiotics are the most common cause of drug fever. Among the antibiotics, b-lactams and sulfonamides are the most common causes of drug fever in the CCU setting. More common causes of fever in the CCU setting are anti-arrhythmics, anti-seizure medications, sulfa-containing loop diuretics (furosemide) or stool softeners (Colace) or tranquilizers, sedatives/sleep medications, antihypertensive medications, and b-blockers. As patients are usually receiving multiple medications, it is not always possible to discontinue an agent likely to be the cause of the drug fever. Often two or three agents have to be discontinued simultaneously. The clinician should discontinue the most likely agent that is not life supporting or essential first, in order properly interpret the decrease in temperature if indeed that was the sensitizing agent responsible for the drug fever. If the agent that is likely to cause the drug fever cannot be discontinued, every attempt should be made to find an equivalent nonallergic substitute, i.e., ethacrinic acid in place of furosemide as a loop diuretic for congestive heart failure and a carbapenem in place of a b-lactam. If the agent responsible for the drug fever is discontinued, temperatures will decrease to near normal/normal within 72 hours. If the temperature does not decrease within 72 hours, then the clinician should discontinue sequentially one drug at a time, those that are likely to be the causes of drug fever. Resolution of drug fever means that not only the temperature returns to normal, but also the leukocytosis decreases and the eosinophils disappear in the differential WBC count (Tables 7 and 8) (1,3,5,10,12–24). IV-Line Infections Any invasive intravascular device may be associated with infection, but central IV lines are the ones most likely to result in IV-line sepsis. Other causes of IV-line sepsis Table 7 Clinical Features of Drug Fever History Many individuals are atopic ‘‘sensitizing medication’’ taken for days or years Physical examination Look ‘‘inappropriately well’’ for degree of fever Low to high-grade fevers (102 F 106 F) Relative bradycardia (with temperature 102 F if not on b-blockers, etc.) No rasha Laboratory tests " WBC count (often with left shift) Eosinophils usually present in peripheral smear (often missed with autonated counters) Eosinophilia is uncommon " ESR in most (may reach 100 mm/hr) Mild/transient "of serum transaminases (early) Abbreviations: WBC, white blood cells; ESR, erythrocyte sedimentation rate. a If present, diagnosis is drug rash with fever. Source: From Refs. 11–15.

58

Cunha

Table 8 Drug Fever (2 Sensitizing Medications) Common causes Antibiotics b-lactams Sleep medications Antiseizure medications Sulfa-containing drugs Stool softeners (Colace) Diuretics (Lasix) Antimicrobials (TMP-SMX, pentamidine) Antidepressants/tranquilizers Antiarrhythmics b-blockers Ace inhibitors

Rare causes Digoxin Steroids Diphenhydramine (Benadryl) Aspirin Vitamins Tetracyclines Erythromycins Ketolides Clindamycin Aminoglycosides Chloramphenicol Vancomycin Teichoplanin Aztreonam Carbapenems Quinolones Quinipristin/dalfopristin Daptomycin Tigecycline

Abbreviation: TMP-SMX, trimethoprim–sulfamethoxazole. Source: From Ref. 12.

that may be encountered in the CCU are an infected Hickman/Broviac, Pik line, or pacemaker lead/generator infection, Quinton catheter. Patients with AV-graft infections resemble, in clinical presentation, those with central IV-line sepsis. The diagnosis of IV-line infection may be obvious or less straightforward. The likelihood that a patient in the CCU has IV-line infection is related to the duration that the central IV line is in place. Central IV-line infections are rare in less than or equal to seven days after line placement. There is progressive increase in the incidence of central IV-line infection following seven days of catheter insertion, i.e., the longer the central IV line is in the more likely that IV sepsis will ensue. Central IV-line infections often present as otherwise unexplained obscure fevers. Half of the patients will have obvious signs of infection at the catheter entry site. This is all that is required for a presumptive diagnosis of IV-line infection, and the catheter should be removed and semiquantitative catheter-tip cultures and blood cultures should be obtained to confirm the diagnosis. The more common problem is in the other half of patients who have no local signs of infection at the site of IV catheter insertion. IV-line infection should be suspected after other diagnostic possibilities have been eliminated in patients who have had a central IV line in place for days/weeks. Blood cultures should be obtained and the catheter removed for semiquantitative culture of the catheter tip. The finding of a positive catheter-tip culture is one with 15 colonies plated in the method of Maki/Cleary. Positive catheter-tip culture without bacteremia indicates only a colonized catheter. Bacteremia without positive catheter-tip culture with the same organism indicates bacteremia but not secondary to the IV line. IV-line infection is diagnosed by demonstrating the same organism in the blood and the catheter tip.

Clinical Approach to Fever in the Critical Care Unit

59

The treatment for central IV-line infection is to remove the central IV line. If no further central venous access is necessary, the line may be discontinued, but if continued central IV line access is required, the catheter may be changed over a guide wire. Changing the catheter over a guide wire does not subject the patient to the possibility of a pneumothorax from a subclavian insertion. Alternately, after the catheter is removed, another catheter may be placed in a different anatomical location. Femoral catheters are the ones most likely to be infected, followed by IJ-inserted catheters. The subclavian inserted central IV lines are those least likely to be infected over time. Central IV-line infections are treated by catheter removal and antibiotics are usually given, even though the source of the bacteremia has been removed. The organisms from the skin, i.e., S. aureus, S. epidermidis/coagulase-negative staphylococci (CoNS), are the most frequent cause, but aerobic gram-negative bacilli and to a lesser extent enterococci are also important causes of IV-line sepsis in the CCU. Many times catheters are often needlessly changed when patients, particularly postoperative patients, spike a fever in the first two to three days postoperatively. Catheter change so early is unnecessary because IV-line infections are rare before being in place at least seven days. If antibiotics are used to treat IV-line infections after the central line is removed, treatment is ordinarily for seven days for gram-negative organisms and for two weeks for gram-positive organisms (excluding CoNS). CoNS are not ordinarily treated because they are low virulence pathogens and are incapable of infection in the absence of prosthetic metal or plastic materials. Even if prosthetic materials are in place in a patient with a CoNS bacteremia, patients have endothelialized their appliances and the likelihood of infection from a transient bacteremia associated with a central line diminishes. It cannot be emphasized too strongly that the clinician should have a high index of suspicion for central IV-line infection the longer the catheter has been in place in patients without an alternate explanation for their prolonged fevers. Catheter lines should not be changed/removed prophylactically if they are in place for less than seven days unless there are obvious signs of infection at the catheter site entry point (1,5,10,40–42). Persistence of Fever The clinical approach to the delayed resolution of fever, persistence of fever, or new appearance of fever related to a complication of therapy, i.e., drug fever after initial improvements in temperature/fever, a recrudescence of fever manifested by new fever/fever spikes may be related to the infectious process, or may be related to a noninfectious complication unrelated to therapy, i.e., myocardial infarction, gastrointestinal hemorrhage, acute pancreatitis, acute gout, deep vein thrombosis, phlebitis, and pulmonary emboli/infarcts. The time that the fever spike occurs in relation to the initial defervescence, pulse temperature relationships, and other associated findings are the key determinants diagnostically in sorting out possible explanations for the reappearance of fever in CCU patients. The recrudescence of fever is virtually never due to resistant organisms. Recrudescence of fever may be due to other infectious processes, i.e., candidemia, invasive aspergillosis, in patients with central lines, or on prolonged/high dose steroid or immunosuppressive therapy. Lack of response to antimicrobial therapy suggests inadequate spectrum or insufficient activity against the pathogen in the antibiotic regimen that is selected (Table 9) (42–44).

60

Cunha

Table 9 Persistent Fever in the Critical Care Unit Antibiotic related problems Inadequate coverage/spectrum Inadequate antibiotic blood levels Inadequate antibiotic tissue levels Undrained abscess Foreign body–related infection Protected focus Abscess Foreign body Chronic osteomyelitis, etc. Organ hypoperfusion/diminished local blood supply In vitro susceptibility but inactive in vivo Antibiotic tolerance (gram-positive cocci) Drug-induced interactions Antibiotic inactivation Antibiotic antagonism Nonantibiotic related problems Treating colonization Noninfectious diseases mimics SLE Drug reactions Drug fever Atelectasis Pleural effusions Seroma Dehydration Acute pancreatitis Pulmonary emboli Acute myocardial infarction CNS hemorrhage/cerebrovascular accident Antibiotic-unresponsive infectious diseases Viral infection Abbreviations: SLE, systemic lupus erythematosus; CNS, central nervous system. Source: From Refs. 1, 3, 5.

CLINICAL DIAGNOSTIC APPROACH TO FEVER IN THE CCU Patients in the CCU with fever are admitted for a primary problem but also arrive with a variety of pre-existing disorders that may interact or complicate the primary reason for admission to the CCU. Problems that occur in the CCU related to new problems, complications of the original/new problems, plus the effect of multiple medications make diagnostic possibilities to explain fever in the CCU complex. The cause of fever may be suggested by epidemiologic factors as well as the history, physical, laboratory, and radiology tests. If the main thrust of the diagnostic approach is to identify reversible/curable causes of fever, analysis of the fever characteristics is the best way to sort out differential diagnostic possibilities in the CCU. Careful attention should be given to whether the fever spike is isolated or sustained, whether the fever is > or 102 F without relative bradycardia, which are sustained. Although there are many possibilities to explain these fevers, i.e., superimposed CMV or bacterial infections, the most important correctable factor to identify as the cause of the fever is inadequate steroid dosage. Patients on chronic corticosteroids when admitted to the CCU should require stress doses of corticosteroids. Without increasing the corticosteroid daily dose, patients develop either a fever from a flare of their SLE/relative bradycardia or adrenal insufficiency, which presents as otherwise unexplained fever in such patients (Table 11) (1,10,47–49). CLINICAL THERAPEUTIC APPROACH General therapeutic interventions should be done as soon as possible. Gastrointestinal hemorrhage may require blood transfusion, collagen vascular diseases/vasculitis may require high dose corticosteroid therapy, pulmonary emboli may require anticoagulation, myocardial infarction may require balloon pump support or cardiac interventional procedures/surgery, IV lines should be removed and sent for semi-quantitative catheter-tip cultures and peripheral blood cultures. Patients with Table 11 Clinical Diagnostic Approach to Fever in Critical Care Unit Early infectious disease consultation All febrile CCU patients should have an infectious disease consultation Infectious disease consultation to evaluate mimics of infections (pseudosepsis) and microbiologic data Persistent low grade fevers (102 F) Noninfectious medical diseases most likely Infectious disease causes also important Acute high, spiking fevers (102 F) Infectious disease etiology most likely Medical disorders excluded by fevers 102 F: MI/LVF PE Acute pancreatitis ARDS Atelectasis/dehydration Hematomas Only noninfectious diseases with temperatures 102 F in CCU Drug fevers Malignant neuroleptic syndrome Central fevers Fevers 2 to blood/blood product transfusion Transient bacteremias 2 to manipulation of a colonized/infected mucosa

Thrombophlebitis C. difficile diarrhea GI hemorrhage Cholecystitis Uncomplicated wound infection

Abbreviations: CCU, critical care unit; ARDS, acute respiratory distress syndrome; GI, gastrointestinal, MI, myocardial infarction; LVF, left ventricular failure; PE, pulmonary emboli. Source: From Refs. 1, 3.

Clinical Approach to Fever in the Critical Care Unit

65

ARDS should be on PEEP with high oxygen concentrations. Abscesses should be drained as soon as possible. Abdominal computed tomography (CT) scanning is invaluable in accessing abdominal pain in a febrile patient in the CCU. Plain films of the abdomen are helpful, and ultrasound is (except for biliary tract obstruction) insufficiently accurate compared to a CT scan. Because CCU problems are time critical, the abdominal CT scan should be obtained on an urgent basis and serially if necessary (Table 12) (50–53). Infectious diseases should be treated with appropriate empiric antimicrobial therapy. Coverage should be directed against the usual pathogen(s) associated with the infected organ system involved. Once again, infectious disease consultation is invaluable in determining adequate and appropriate antibiotic therapy without being excessive. Infectious disease consultants can also streamline the antibiotic regimen and make adjustments for drug allergies as well as hepatic and/or renal insufficiency, and take into account significant drug interactions/side effects to tailor the antimicrobial therapy to the patient’s condition. The antibiotic selected should have a spectrum appropriate for the site of infection, be started as soon as possible, have a low resistance potential, have excellent safety profile, and be inexpensive. Coverage should be adequate for likely pathogens, but colonization should not be treated. The infectious disease consultant is in the best position to analyze complex/

Table 12 Therapeutic Approach to Fever in the Critical Care Unit Microbiologic data evaluation Critical to differentiate colonization from infection: Respiratory secretion isolates Urinary isolates Analysis of origin of blood culture isolates Rule out pseudo-infections Common causes of fevers Nosocomial pneumonia/VAP Chest X ray If negative, no nosocomial pneumonia/VAP If positive, rule out LVF, ARDS, etc. Perform semiquantitative BAL to confirm diagnosis Check central IV lines Duration of insertion The longer the IV line is overdue, the more likely the fever is due to IV-line infection Otherwise unexplained fevers in a patient with overdue IV lines should be regarded as IV line infection until proven otherwise Evidence of infections at local site If IV shows sign of infection, remove IV line immediately, send tip for semiquantitative culture, and obtain blood cultures from peripheral vein If IV site non-erythematous, IV line infection not ruled out, remove/replace IV line and send removed catheter tip for semiquantitative culture If nosocomial pneumonia and IV-line infection eliminated from diagnostic consideration, consider drug fever Early empiric therapy Coverage based on site/organism correlations: colonization should not be treated Infectious disease consultant recommendations should be followed. Abbreviations: IV, intravenous; ARDS, acute respiratory distress syndrome; BAL, bronchoalveolar lavage; VAP, ventilation associated pneumonia; LVF, left ventricular failure. Source: From Refs. 1, 3, 5.

S. pneumoniae, H. influenzae, K. pneumoniae

Group D streptococcib E. faecalis E. faecium (VRE)

Enterobacteriaceae B. fragilis

Usual pathogens

S. epidermidis, IV-line sepsis S. aureus (MSSA), Bacterial (Treat Klebsiella, initially for MSSA; Enterobacter, if later identified as Serratia MRSA, treat accordingly)

Lung source

Unknown source

Subset a

Alternate IV therapy

IV-to-PO switch

Quinolone (IV)  2 wk plus either Moxifloxacin 400 mg (PO) Meropenem 1 g (IV) q8h  2 wk or q24h  2 wk or metronidazole 1 g (IV) q24h  2 wk piperacillin/tazobactam 4.5 g (IV) combination therapy with or clindamycin 600 mg (IV) q8h  2 wk or imipenem 500 mg clindamycin 300 mg (PO) q8h  2 wk (IV) q6h  2 wk or ertapenem 1 g q8h  2 wk plus either (IV) q24h  2 wk or combination ciprofloxacin 500 mg (PO) therapy with ceftriaxone 1 g (IV) q12h  2 wk or gatifloxacin q24h  2 wk plus metronidazole 1 g 400 mg (PO) q24h  2 wk (IV) q24h  2 wk or levofloxacin 500 mg (PO) q24h  2 wk Ampicillin/sulbactam 3 g (IV) Meropenem 1 g (IV) q8h  2 wk or Quinolonec (PO)  2 wk q6h  2 wk or Quinolonec piperacillin/tazobactam 4.5 g (IV) q8h  2 wk (IV)  2 wk Linezolid 600 mg (PO) Linezolid 600 mg (IV) q12h  2 wk or Chloramphenicol 500 mg (IV) q12h  2 wk or doxycycline q6h  2 wk or Doxycycline 200 mg quinupristin/dalfopristin 7.5 mg/ 100 mg (PO) q12h  2 wk (IV) q12h  3 days, then 100 mg kg (IV) q8h  2 wk (IV) q12h  11 days Ceftriaxone 1 g (IV) q24h  2 wk or Quinolonec (IV) q24h  2 wk or any Quinolonec (PO) q24h  2 wk cefepime 2 g (IV) q12h  2 wk second generation cephalosporin or doxycycline 200 mg (IV)  2 wk (PO) q12h  3 days, then 100 mg (PO) q12h  11 days Meropenem 1 g (IV) q8h  2 wk or Ceftriaxone 1 g (IV) q24h  2 wk or Quinoloned (PO) cefepime 2 g (IV) q12h  2 wk quinoloned (IV) q24h  2 wk q24h  2 wk or cephalexin 500 mg (PO) q6h  2 wk

Preferred IV therapy

Table 13 Enzyme Antibiotic Therapy in Sepsis/Septic Shock

66 Cunha

Aspergillus

Non-albicans Candidae

Fungal (Treat initially Candida albicans for non-albicans Candida; if later identified as C. albicans, treat accordingly)

S. aureus (MRSA)

(Continued )

Vancomycin 1 g (IV) q12h  2 wk or Linezolid 600 mg (PO) q12h  2 wk or linezolid 600 mg (IV) q12h  2 wk minocycline 100 mg (PO) or quinupristin/dalfopristin q12h  2 wk 7.5 mg/kg (IV) q8h  2 wk Fluconazole 400 mg (PO) Preferred IV therapy: fluconazole Preferred IV therapy: fluconazole q24h  2 wk or 800 mg (IV)  1, then 400 mg (IV) 800 mg (IV)  1, then 400 mg (IV) voriconazole (see ‘‘usual q24h  2 wk q24h  2 wk dose,’’ p. 480) Alternate IV therapy: or caspofungin Itraconazole 200 mg (PO) Alternate IV therapy: or solution q12h  2 wk 70 mg (IV)  1 dose, then 50 mg Caspofungin 70 mg (IV)  1 dose, (IV) q24h  2 wk or lipidthen 50 mg (IV) q24h  2 wk or associated formulation of lipid-associated formulation of amphotericin B (p. 369) (IV) amphotericin B (p. 369) (IV) q24h  2 wk or amphotericin B q24h  2 wk or amphotericin B deoxycholate 0.7 mg/kg (IV) deoxycholate 0.7 mg/kg (IV) q24h  2 wk or voriconazole (see q24h  2 wk or voriconazole (see ‘‘usual dose,’’ p. 480) or ‘‘usual dose,’’ p. 480) or itraconazole 200 mg (IV) q12h  2 itraconazole 200 mg (IV) q12h  2 days, then 200 mg (IV) q24h  2 wk days, then 200 mg (IV) q24h  2 wk Caspofungin or lipid amphotericin B Fluconazole or itraconazole (see Fluconazole (see C. albicans, or amphotericin B deoxycholate C. albicans, above) or voriconazole above) or itraconazole (see C. albicans, above)  2 wk (see ‘‘usual dose,’’ p. 480)  2 wk 200 mg (PO) solution q24h or voriconazole (see ‘‘usual dose,’’ p. 480)  2 wke Voriconazole (see ‘‘usual Lipid-associated formulation of Voriconazole (see ‘‘usual dose’’, p. dose,’’ p. 480)  2 wk or amphotericin B (p. 369) (IV) 480)  2 wk or caspofungin 70 mg itraconazole 200 mg (PO) q24h  2 wk or amphotericin B (IV)  1 dose, then 50 mg (IV) solution q12h  2 days, deoxycholate 1–1.5 mg/kg (IV) q24h  2 wk or itraconazole then 200 mg (PO) solution q24h  2 wk 200 mg (IV) q12h  2 days, then q24h  2 wk 200 mg (IV) q24h  2 wk Daptomycin 6 g/kg (IV) 24 h or linezolid 600 mg (IV) q12h  2 wk

Clinical Approach to Fever in the Critical Care Unit 67

Enterobacteriaceae

Urosepsis Gram () bacilli

Gram þ streptococci Group B streptococci E. faecalis (treat initially for E. faecalis; if later identified as VRE, treat accordingly) Treat initially for E. faecium (VRE) E. faecalis; if later identified as VRE, treat accordingly) Organism not known Enterobacteriaceae Group B, D streptococci Aspelnia or S. pneumoniae, hyposplenia H. influenzae, N. meningitidis

Enterobacteriaceae B. fragilis

Usual pathogens

Cefepime 2 g (IV) q12h  2 wk or Cefotaxime 2 g (IV) q6h  2 wk

Ceftriaxone 2 g (IV) q24h  2 wk or Quinolonef (IV) q24h  2 wk

Quinolonef (PO) q24h  2 wk

Linezolid 600 mg (PO) q12h  1–2 wk or doxycycline 200 mg (PO) q12h Quinolone (POa)  2 wk

Linezolid 600 mg (IV) q12h  1–2 wk or quinupristin/ dalfopristin 7.5 mg/kg (IV) q8h  1–2 wk Piperacillin/tazobactam 4.5 g (IV) q8h  1–2 wk

Quinolone (IV)a  1–2 wk

Ceftriaxone 1 g (IV) q24h  1–2 wk

Linezolid 600 mg (IV) q12h  1–2 wk or quinupristin/ dalfopristin 7.5 mg/kg (IV) q8h  1–2 wk Quinolone (IV)a  2 wk

IV-to-PO switch

Quinolonea (IV)  2 wk plus either Moxifloxacin 400 mg (PO) q24h  2 wk or metronidazole 1 g (IV) q24h  2 wk combination therapy with or clindamycin 600 mg (IV) clindamycin 300 mg (PO) q8h  2 wk q8h  2 wk plus either ciprofloxacin 500 mg (PO) q12h  2 wk or gatifloxacin 400 mg (PO) q24h  2 wk or levofloxacin 500 mg (PO) q24h  2 wk Aztreonam 2 g (IV) q8h  1–2 wk or Quinolone (PO)a  1–2 wk Any aminoglycoside (IV)  1–2 wk Amoxicillin 1 g (PO) Ampicillin 1–2 g (IV) q4h  1–2 wk q8h  1–2 wk or quinolone or vancomycin 1 g (IV) q12h  (PO)a  1–2 wk 1–2 wk

Alternate IV therapy

Meropenem 1 g (IV) q8h  2 wk or piperacillin/tazobactam 4.5 g therapy with (IV) q8h  2 wk or tigacycline 100 mg (IV)  1 dose then 50 mg (IV) of 12 h2 wk or combination therapy with ceftriaxone 1 g (IV) of 24h  2 wk plus metronidazole 1 g (IV) q24h  2 wk

Preferred IV therapy

Enzyme Antibiotic Therapy in Sepsis/Septic Shock (Continued )

Intra-abdominal/ pelvic source

Subset

Table 13

68 Cunha

Gram-negative or gram-positive bacteria

Severe sepsis

Treat the same as for fungal infection Treat the same as for fungal infection Treat the same as for fungal (pp. 118–119) (pp. 118–119) infection (pp. 118–119) Treat the same as pulmonary TB Treat the same as pulmonary TB Treat the same as pulmonary (p. 47) plus steroids  1–2 wk (p. 47) plus steroids  1–2 wk TB (p. 47) plus steroids  1–2 wk Treat with 4 anti-TB drugs Treat with 4 anti-TB drugs (NIH, Treat with 4 anti-TB drugs (NIH, (NIH, rifampin, rifampin, ethambutol, cycloserine) rifampin, ethambutol, cycloserine) ethambutol, cycloserine) q24h  6–12 months plus steroids q24h  6–12 months plus steroids q24h  6–12 months plus (e.g., prednisone 40 mg q24h)  (e.g., prednisone 40 mg q24h)  steroids (e.g., prednisone 1–2 wk 1–2 wk 40 mg q24h)  1–2 wk Appropriate antimicrobial therapy  drotrecogin alpha (Xigris) 24 mg/ Appropriate antimicrobial plus surgical decompression/ kg/hr therapy plus surgical drainage if needed  decompression/drainage if needed plus drotrecogin alpha (Xigris) 24 mg/kg/hr

Note: Duration of therapy represents total time IV or IV þ PO. Loading dose not needed PO if given IV with the same drug. Most patients on IV therapy able to take PO meds should be switched to PO therapy after clinical improvement. a Ciprofloxacin 400 mg (IV) or 500 mg (PO) q12h or gatifloxacin 400 mg (IV or PO) q24h or levofloxacin 500 mg (IV or PO) q24h. b Treat initially for E. faecalis; if later identified as E. faecium (VRE), treat accordingly. c Ciprofloxacin 400 mg (IV) or 500 mg (PO) q12h or gatifloxacin 400 mg (IV or PO) q24h or levofloxacin 500 mg (IV or PO) q24h or moxifloxacin 400 mg (IV or PO) q24h. d Gatifloxacin 400 mg or levofloxacin 500 mg or moxifloxacin 400 mg. e Best agent depends on infecting species. Fluconazole-susceptibility varies predictably by species. C. glabrata (usually) and C. krusei (almost always) are resistant to fluconazole. C. lusitaniae is often resistant to amphotericin B (deoxycholate and lipid-associated formulations). Others are generally susceptible to all agents. f Gatifloxacin 400 mg or levofloxacin 500 mg or moxifloxacin 400 mg. Abbreviations: IV, intravenous; BCG, bacillus Calmette–Guerin; MSSA/MRSA, methicillin-sensitive/resistant S. aureus; TB, tuberculosis; VRE, vancomycin resistant enterococci; INH, isoniazid. Source: From Ref. 53.

BCG

M. tuberculosis

Candida, Aspergillus

Miliary BCG (disseminated)

Steroids (high chronic dose) Miliary TB

Clinical Approach to Fever in the Critical Care Unit 69

70

Cunha

Table 14 Treatment of Fever in the Critical Care Unit Obligatory Reduction in Temperatures >106 F Heat stroke Malignant hyperthermia Malignant neoplastic syndrome Central fevers Drug fever Obligatory reduction in temperatures >102 Fa Acute myocardial infarction Borderline pulmonary function CNS trauma Optional Reduction of Temperatures >102 Fa Blood/blood product transfusion reactions Postdiagnosis of infectious/noninfectious diseases febrile disorders a Temperatures should be lowered to 106 F should be treated to reduce the temperature to 102 F to 104 F range. Patients with CNS trauma, recent myocardial infarction, or borderline cardiopulmonary function should have temperatures maintained at 102 F. Temperatures >102 F in sick patients could worsen neurologic daze in those with CNS trauma or precipitate an acute myocardial infarction, chronic heart failure, or pulmonary decompensation in those with advanced cardiopulmonary disease (Table 14) (54–56). REFERENCES 1. Cunha BA. Clinical approach to fever in the Critical Care Unit. Crit Care Clin 1998; 8: 1–14. 2. Clarke DE, Kimelman J, Raffin TA. The evaluation of fever in the intensive care unit. Chest 1991; 100:213–220. 3. Cunha BA. Fever in the intensive care unit. Intensive Care 1999; 25:648–651. 4. Marik PE. Fever in the ICU. Chest 2000; 117:855–869. 5. Cunha BA, Shea KW. Fever in the intensive care unit. Infect Dis Clin North Am 1996; 10:185–209.

Clinical Approach to Fever in the Critical Care Unit

71

6. McGowan JE, Rose RC, Jacobs NF, et al. Fever in hospitalized patients. Am J Med 1987; 82:580–586. 7. O’Grady NP, Barie PS, Bartlett J, et al. Practice parameters for evaluating new fever in critically ill patients. Task Force of the American College of Critical Care Medicine of the Society of Critical Care Medicine in collaboration with the Infectious Disease Society of America. Crit Care Med 1998; 26:392–408. 8. Stumacher RJ. Fever in the ICU. Infect Dis Pract 1996; 20:89–92. 9. Ryan M, Levy MM. Clinical review: fever in intensive care unit patients. Crit Care 2003; 7:221–225. 10. Cunha BA. Approach to fever. In: Gorbach SL, Bartlett JB, Blacklow NR, eds. Infectious Diseases in Medicine Surgery. 4th ed. Baltimore: Lippincott Williams & Wilkins, 2005: 54–63. 11. Cunha BA. The diagnostic significance of fever curves. Infect Dis Clin North Am 1996; 10:33–44. 12. Johnson DH, Cunha BA. Drug fever. Infect Dis Clin North Am 1996; 10:85–91. 13. Mackowiak PA, LeMaistre CF. Drug fever: a critical appraisal of conventional concepts. Ann Intern Med 1987; 106:728–733. 14. Mackowiak PA. Drug fever: mechanisms, maxims and misconceptions. Am J Med Sci 1987; 294:275–286. 15. Cunha BA. The diagnostic approach to rash and fever in the critical care unit. Crit Care Clin 1998; 8:35–54. 16. Cunha BA. Rash and fever in the intensive care unit. In: Abraham E, Vincent JL, Kochanek P, eds. Textbook of Critical Care Medicine. 5th ed. Philadelphia: Elsevier, 2005:113–119. 17. Cunha BA. Fever in malignant disorders. Infect Dis Pract 2004; 28:335–336. 18. Ginsberg MD, Busto R. Combating hyperthermia in acute stroke: a significant clinical concern. Stroke 1998; 29:529–534. 19. Wood AJJ. Adverse drug reactions. In. Fauci AS, Braunwald E, Isselbacher KJ, et al., eds. Harrison’s Principles of Internal Medicine. New York: McGraw–Hill, 1998:422–430. 20. Ferrara JL. The febrile platelet transfusion reaction: a cytokine shower. Transfusion 1995; 35:89–90. 21. Fenwick JC, Cameron M, Naiman SC, et al. Blood transfusion as a cause of leucocytosis in critically ill patients. Lancet 1994; 344:855–856. 22. Marx N, Neumann FJ, Ott I, et al. Induction of cytokine expression in leukocytes in acute myocardial infarction. J Am Coll Cardiol 1997; 30:165–170. 23. Murray HW, Ellis GC, Blumenthal DS, et al. Fever and pulmonary thromboembolism. Am J Med 1979; 67:232–235. 24. Engoren M. Lack of association between atelectasis and fever. Chest l995; 107:81–84. 25. Cunha BA. FUO in the ICU. In Bouza E, Picazo JJ, eds. Infeccion. Madrid, Spain, 2004. 26. Cunha BA. The diagnostic significance of relative bradycardia in infectious disease. Clin Microbiol Infect 2000; 6:633–634. 27. Cotton LM, Strampfer MJ, Cunha BA. Legionella and mycoplasma pneumonia: a community hospital experience. Clin Chest Med 1987; 8:441–453. 28. Ostergaard L, Huniche B, Andersen PL. Relative bradycardia in infectious diseases. J Infect 1996; 33:185–191. 29. Cunha BA. Diagnostic significance of relative bradycardia. Infect Dis Pract 1997; 21: 38–40. 30. Cunha BA. Relative bradycardia as a diagnostic clue. Intern Med 1999; 20:42–46. 31. Wittesjo B, Bjornham A, Eitrem R. Relative bradycardia in infectious diseases. J Infect 2000; 38:246–247. 32. Meduri GU, Mauldin GI, Wunderink RG, et al. Causes of fever and pulmonary densities in patients with clinical manifestations of ventilator-associated pneumonia. Chest 1994; 106:221–235.

72

Cunha

33. Meduri GU. Diagnosis and differential diagnosis of ventilator-associated pneumonia. Clin Chest Med 1995; 16:61–93. 34. Santos E, Talusan, Brandstetter RD. Roentgenographic mimics of pneumonia in the critical care unit. Crit Care Clin 1998; 14:91–104. 35. Cunha BA. Severe community-acquired pneumonia. Crit Care Clin 1998; 14:105–117. 36. Cunha BA. Nosocomial pneumonia: diagnostic and therapeutic considerations. Med Clin North Am 2001; 85:79–114. 37. Cunha BA. Pneumonia Essentials. Michigan: Physicians Press, 2006. 38. Eisenstein L, Cunha BA. Herpes simples virus type I (HSV-I) pneumonia presenting as failure to wean. Heart Lung 2003; 32:65–66. 39. Cunha BA. Herpes simplex-1 (HSV-1) pneumonia. Infect Dis Pract 2005; 29:375–378. 40. Maki DG. Pathogenesis, prevention, and management of infections due to intravascular devices used for infusion therapy. In: Bisno A, Waldvogel F, eds. Infections Associated with Indwelling Medical Devices. Washington, D.C.: American Society for Microbiology, 1989:161–177. 41. Cunha BA. Intravenous line infections. Crit Care Clin 1998; 8:339–346. 42. Garibaldi RA, Brodine S, Matsumiya S, et al. Evidence for the noninfectious etiology of early postoperative fever. Infect Control 1985; 6:273–277. 43. Kitaichi M. Differential diagnosis of bronchilitis obliterans organizing pneumonia. Chest 1992; 102:44–49. 44. Fry DE. Postoperative fever. In: Mackowiak PA, ed. Fever: Basic Mechanisms and Management. New York: Raven Press, 1991:243–254. 45. Warren JW. Catheter-associated urinary tract infections. Infect Dis Clin North Am 1997; 11:609–619. 46. Paradisi F, Corti G, Mangani V. Urosepsis in the critical care unit. Crit Care Clin 1998; 114:165–180. 47. Tu RP, Cunha BA. Significance of fever in the neurosurgical intensive care unit. Heart Lung 1988; 17:608–611. 48. Bouza E, Munoz P, Alonso R. Clinical manifestations, treatment and control of infections caused by Clostridium difficile. Clin Microbiol Infect 2005; 4:57–64. 49. Caines C, Gill MV, Cunha BA. Non-Clostridium difficile nosocomial diarrhea in the intensive care unit. Heart Lung 1997; 26:83–84. 50. Mieszczanska H, Lazar J, Cunha BA. Cardiovascular manifestations of sepsis. Infect Dis Pract 2003; 27:183–186. 51. Brun–Buisson C, Doyon F, Carlet J, et al. Incidence, risk factors, and outcome of severe sepsis and septic shock in adults: a multicenter prospective study in intensive care units. French ICU Group for Severe Sepsis. JAMA 1995; 274:968–974. 52. Cunha BA. Antibiotic treatment of sepsis. Med Clin North Am 1995; 79:551–558. 53. Cunha BA. Antibiotic Essentials. (5th Ed) Michigan: Physicians Press, 2006. 54. Plaisance KI, Mackowiak PA. Antipyretic therapy: physiologic rationale, diagnostic implications, and clinical consequences. Arch Intern Med 2000; 160:449–456. 55. Mackowiak PA. Physiological rationale for suppression of fever. Clin Infect Dis 2000; 31(suppl 5):S185–S189. 56. Cunha BA. Should fever be treated in sepsis. In: Vincent JL, Carlet J, Opal S, eds. Sepsis. 21st Symposium on Intensive Care and Emergency Medicine. New York: Kluwer Academic Publishers, 2001:705–717.

4 Sepsis and Its Mimics in the Critical Care Unit Burke A. Cunha Infectious Disease Division, Winthrop-University Hospital, Mineola, and State University of New York School of Medicine, Stony Brook, New York, U.S.A.

INTRODUCTION Sepsis refers to bacteremia or fungemia with hypotension and organ dysfunction. The main clinical problem with the ‘‘septic’’ patient is to determine if the patient is septic or has a noninfectious condition that mimics sepsis by hemodynamic or laboratory parameters. In the intensive care setting it is of critical importance to differentiate between sepsis and its mimics (1–6). Diagnostic Approach Many patients with fever and hypotension are not septic. Several clinical disorders resemble sepsis. Patients do not become septic without a major breach in host defenses. The most important clinical consideration in determining whether a patient is septic is to identify the source of infection. Sepsis is a complication with only relatively few infections. Infections limited to specific infections in a few organ systems are the only ones with septic potential. Most sepsis is derived from perforated obstructions or abscesses of the gastrointestinal (GI) tract/pelvis, hepatobiliary tract, and genitourinary (GU) tract, or may be related to central intravenous (IV) lines. Even though the GI tract is the most frequent focus of infection leading to sepsis, not all GI disorders including infections have a septic potential. Lower GI tract perforations, intra-abdominal/pelvic abscesses, and pylephlebitis commonly result clinically in sepsis. In contrast, gastritis and nonperforating gastric ulcer are rarely associated with sepsis. Cholangitis in the hepatobiliary tract results in sepsis, but rarely, if ever, complicates acute/chronic cholecystitis (6–13). IV line sepsis represents the ultimate breach in host defenses, as the pathogenic organisms from central catheters are introduced directly into the bloodstream in high concentrations (14,15). The primary task is to search for GI, GU, or an IV source of sepsis. It is almost always possible to identify the septic source by physical exam, laboratory, or radiology tests. Without local signs of entry site infection, IV-line sepsis should not be entertained if the central IV line has been in place less than seven days. 73

74

Cunha

If intraabdominal and GU sources have been eliminated as diagnostic possibilities, central IV lines, either temporary or long term, should be considered as a cause of sepsis. The longer a central IV line is in place, the more likely the central IV line may be the cause of fever/hypotension. Signs of infection at entry sites of central IV lines indicate likely IV-line sepsis, but no superficial erythema/swelling does not rule out IV-line sepsis (14–16). Disorders that mimic sepsis should be recognized to treat the condition and not to avoid inappropriate treatment with antibiotics. Disorders that mimic sepsis (pseudosepsis) include GI hemorrhage, pulmonary embolism, myocardial infarction, acute pancreatitis, diabetic ketoacidosis, systemic lupus erythematosus (SLE) flare, and relative adrenal insufficiency, inadequate (maintenance, not stress dosed), or too rapidly typed steroid therapy, ventricular pseudo-aneurysm, massive aspiration or atelectasis, systemic vasculitis, and diuretic-induced hypovolemia (Table 1) (6,17–21). Clinical Signs of Sepsis Excluding the elderly, compromised hosts, and uremic patients, fever is a cardinal sign of inflammation or infection. Fever should not be equated with infection as the chemical mediators of inflammation and infection, i.e., cytokines, induce a febrile response mediated via the preoptic nucleus of the anterior hypothalamus. All that is febrile is not infectious, and most, but not all diseases causing sepsis are accompanied by temperatures 102 F. With the exceptions of drug fever and adrenal insufficiency, the disorders that mimic sepsis and pseudosepsis have temperatures 102 F. The temperature relationships are critical when considered together with organ involvement, i.e., GI, GU, etc. are key factors in determining if the patient is septic or has a noninfectious disorder resembling sepsis. Hyperthermia 106 F is only caused by noninfectious disorders. Hypothermia is an important clinical clue to bacteremia, particularly in renal insufficiency. In normal hosts with fever, sepsis should not be a diagnostic consideration if temperatures are 106 F (Table 2) (22–25). LABORATORY ABNORMALITIES IN SEPSIS The usual hemodynamic parameters associated with sepsis include decreased peripheral resistance (PR) with increased cardiac output (CO) accompanied by tachycardia/respiratory alkalosis. Patients with fever are often diagnosed as septic. Although sepsis is associated with hemodynamic abnormalities, i.e., # PR/" CO, many disorders mimicking sepsis also have similar findings, i.e., acute pancreatitis, GI bleed, etc. If hemodynamic abnormalities are present but are not accompanied by GI, GU, or IV clinical disorders associated with sepsis, then it should be assumed that the patient has a noninfectious mimic of sepsis. As with hemodynamic parameters, laboratory data may mislead the unwary in incorrectly ascribing laboratory abnormalities to an infectious rather than a noninfectious process. An increase in white peripheral blood cell count with a shift to the left is a nonspecific reaction to stress, and is not specific for infection. Leukocytosis does not differentiate bacterial from viral infections. An increase in white count with a shift to the left is a measure of the intensity of the systemic response to stress of infectious or noninfectious disorders. Similarly, an increase in fibrin split products and lactic acid, decrease in serum albumin, a-2 globulins, and fibrinogen, or an increase in prothrombin time/partial thromboplastin time are compatible but not characteristic of infection.

Sepsis and Its Mimics in the Critical Care Unit

75

Table 1 Clinical Conditions Associated with Sepsis Disorders Associated with sepsis (fevers  102 F) GI source Liver Abscess Gallbladder Gallbladder ‘‘wall abscess’’ Cholangitis Colon Colitis Diverticulitis Toxic megacolon Perforation Obstruction Abscess Genitourinary source Renal Pyelonephritis Intra/perinephric abscess Calculi Urinary tract obstruction Partial Total Prostate Abscess Pelvic source Pelvic peritonitis Tubo-ovarian abscess Pelvic septic thrombophlebitis Lower respiratory source CAP Asplenia/hyposplenism Empyema Lung abscess Nosocomial pneumonia Intravascular source IV line infection Central IV lines PICC lines Hickman/Broviac catheters Infected prosthetic devices AV grafts Jugular vein septic thrombophlebitis Cardiovascular source Acute bacterial endocarditis Myocardial abscess Paravalvular abscess Other Toxic shock syndrome

Not associated with sepsis (fevers  102 F) GI source Esophagitis Gastritis Pancreatitis GI bleed Genitourinary source Urethritis Cystitis (normal hosts) Cervicitis Vaginitis PID Catheter-associated bacteriuria (normal hosts) Upper respiratory source Pharyngitis Sinusitis Mastoiditis Bronchitis Otitis Lower respiratory source CAP (normal host) Skin/soft tissue source Osteomyelitis Uncomplicated wound infections Cardiovascular source Subacute bacterial endocarditis Central nervous system source Bacterial meningitis (excluding meningococcal meningitis with meningococcemia) Intravascular source A-lines Peripheral IV lines

Abbreviations: GI, gastrointestinal; IV, intravenous; PID, pelvic inflammatory disease; CAP, community acquired pneumonia; BPH, benign prostatic hyperpertrophy PICC, peripherally inserted central catheter; AV, arteriovenous. Source: From Refs. 9, 22.

76

Cunha

Table 2 Clinical Mimics of Sepsis Acute gastrointestinal hemorrhage Acute pulmonary embolism Acute myocardial infarction Acute pancreatitis Diabetic ketoacidosis SLE flare Relative adrenal insufficiency Diuretic-induced hypovolemia Rectus sheath hematoma Source: From Refs. 9, 22.

Laboratory parameters that are more indicative of infection include leukopenia or thrombocytopenia. The only laboratory abnormalities that are specific for sepsis are organisms in the blood, i.e., gram/acridine orange stains of buffy-coat smears/high grade positivity in blood cultures (excluding contaminants). Increased cytokine/endotoxin levels are also suggestive. Highly elevated C-reactive protein (CRP) levels have also been described as a marker for sepsis. Positive buffy-coat smears are not present in all patients with bacteremia, but when positive are diagnostic and rapid. The bacteria/fungi present in buffy-coat smears are helpful in determining the origin of the septic process by their association with particular organ system involvement, i.e., poorly stained pleomorphic gram-negative bacilli (Bacteroides fragilis) Table 3 Sepsis vs. Mimics of Sepsis Parameters Microbiologic

Hemodynamic Laboratory

Clinical

Disorders mimicking sepsis Negative blood cultures (excluding skin contaminants)

# PVR " CO " WBC (with left shift) Normal platelet count # Albumin " FSP " Lactate " D-dimers " PT/PTT # Fibrinogen # a2 globulins 102 F Hypotension Tachycardia Respiratory alkalosis

Sepsis (bacteremia from GI/pelvic GU, IV source) Positive buffy-coat smear Bacteremia (excluding skin contaminants) # PVR " CO " WBC (with left shift) # Platelet count # Albumin " FSP " Lactate " D-dimers " PT/PTT

102 F Hypotension Tachycardia Respiratory alkalosis

Abbreviations: PVR, peripheral vascular resistance; CO, cardiac output; FSP, fibrin split products; WBC, white blood cell; PT/PTT, prothrombin time/partial thromboplastin time; GI, gastrointestinal; GU, genitourinary; IV, intravenous. Source: From Refs. 9, 22.

Sepsis and Its Mimics in the Critical Care Unit

77

point to a GI, but not GU/IV source. The morphology/arrangement of the bacteria in buffy-coat smears is also useful in selecting appropriate empiric antibiotic coverage (Table 3) (26–30). EMPIRIC ANTIMICROBIAL THERAPY The selection of appropriate antibiotic therapy for sepsis depends on accurate localization of the infectious process to the abdomen/pelvis, GU tract, or IV line. Because each organ has its normal resident flora that become the pathogenic flora when the organ function is disrupted, empiric coverage is directed against the normal resident flora (Table 4). Factors in antibiotic selection include hepatic/renal insufficiency, allergic status of the patient, tissue penetration of the antibiotic, safety profile of the antibiotic, resistance potential of the antibiotic, and cost. If the spectrum is appropriate for the source of sepsis, no regimen is superior to others in terms of clinical outcome. However, clinicians should utilize the most Table 4 Empiric Therapy of Sepsis Based on Organ System Involved Empiric therapy usual organisms Source/usual organisms Lower GI tract/pelvis (common coliforms plus Bacteroides fragilis) GU tract/kidneys/prostate (aerobic gram-negative bacilli) (Enterococci non-VRE)

(Enterococci VRE)

Organism unknown

Bloodstream IV line (aerobic gram-negative bacilli, Staphylococcus aureus, Enterococci) Lung nosocomial pneumonia/ vent-associated pneumonia (aerobic gram-negative bacilli) a

Monotherapy

Combination therapy Aztreonam or aminoglycoside plus either clindamycin or metronidazole

Meropenem tigacycline Ertapenem Piperacillin/tazobactam Quinolone Third-generation cephalosporin Aztreonam Ampicillin Piperacillin Meropenem Linezolid Daptomycin Quinupristin-dalfopristin Meropenem Piperacillin/tazobactam tigacycline Meropenema

Cefepime plus vancomycin

Meropenem Cefepime Cefoperazone Levofloxacin

Meropenem or cefepime plus either levofloxacin or aztreonam or amikacin

Vancomycin, daptomycin, or linezolid if most intravenous-line infections in institution due to methicillinresistant Staphylococcus aureus (MRSA). Abbreviations: IV, intravenous; GI, gastrointestinal; GU, genitourinary; VRE, vancomycin-resistant enterococci. Source: From Ref. 32.

78

Cunha

clinically/cost-effective regimens with a low resistance potential and begin therapy as soon as the diagnosis of sepsis is made. The basis of empiric therapy for sepsis depends on eliminating the source of sepsis and covering the patient with antibiotic therapy appropriate for the septic source. The use of steroids and anti-cytokine therapies remains controversial and is of unproven benefit (31–44).

SUMMARY The immediate task of the clinician is to determine whether the patient has sepsis or a mimic of sepsis. The diagnostic process may be approached from the negative perspective, i.e., if the patient does not have a GI, GU, and IV process usually associated with sepsis, then the patient in all probability does not have sepsis, and the workup should be directed to diagnosed disorders that mimic sepsis. The temperature of the patient is of key importance in determining if the patient has sepsis or a noninfectious mimic. In temperatures 106 F and 102 F, a noninfectious disease process is likely and argues against a diagnosis of sepsis. Antibiotic therapy should be instituted as soon as there is a basis for the diagnosis of sepsis, i.e., characteristic (perforation, obstruction, or abscess) organ system of infection, GI, GU, and IV site. Coverage should be based on the usual pathogens associated with the involved organ system. Antibiotics with appropriate spectrum, good safety profile, low resistance potential, and anti-endotoxin qualities are preferred. In sepsis related to perforation, obstruction, or abscess, surgical intervention is paramount and should be done as soon as the diagnosis is confirmed.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

14.

Annane D, Bellisant E, Cavaillon JM. Septic shock. Lancet 2005; 365:63–78. Hardaway RM. A review of septic shock. Am Surg 2000; 66:22–29. Murray MJ, Kumar M. Sepsis and septic shock. Postgrad Med 1991; 90:199–202. Sibbald WJ, Marshall J, Christou N, et al. ‘‘Sepsis’’—clarity of existing terminology or more confusion? Crit Care Med 1991; 19:996–998. Sharma S, Kumar A. Septic shock, multiple organ failure, and acute respiratory distress syndrome. Curr Opin Pulm Med 2003; 9:199–209. Cunha BA. Sepsis and its mimics. Intern Med 1992; 13:48–55. Lazaron V, Barke RA. Gram-negative bacterial sepsis and the sepsis syndrome. Urol Clin North Am 1999; 26:687–699. Cunha BA. Urosepsis. J Crit Illness 1997; 12:616–625. Marshall JC. Intra-abdominal infections. Microbes Infect 2004; 6:1015–1025. Sacks-Berg A, Calubiran OV, Cunha BA. Sepsis associated with transhepatic cholangiography. J Hosp Infect 1992; 20:43–50. Carpenter HA. Bacterial and parasitic cholangitis. Mayo Clin Proc 1998; 73:473–478. Alberti C, Brun-Buisson C. The sources of sepsis. In: Vincent JL, Carlet J, Opal SM, eds. The Sepsis Text. Boston: Kluwer Academic Publishers, 2002:491–503. Cruz K, Dellinger RP. Diagnosis and source of sepsis: the utility of clinical findings. In: Vincent JL, Carlet J, Opal SM, eds. The Sepsis Text. Boston: Kluwer Academic Publishers, 2002:11–28. Bouza E, Burillo A, Munoz P. Catheter-related infections: diagnosis and intravascular treatment. Clin Microbiol Infect 2002; 8:265–274.

Sepsis and Its Mimics in the Critical Care Unit

79

15. Gill MV, Cunha BA. IV line sepsis. In: Cunha BA, ed. Infectious Diseases in Critical Care Medicine. New York: Marcel Dekker, 1998:57–65. 16. Cunha BA. Intravenous line infections. Crit Care Clin 1998:339–346. 17. Hamid N, Spadafora P, Khalife ME, Cunha BA: Pseudosepsis: rectus sheath hematoma mimicing septic shock. Heart & Lung 2006; 35:528–530.. 18. McCriskin JW, Baisden CE, Spaccevento LJ, et al. Pseudosepsis after myocardial infarction. Am J Med 1987; 83:577–580. 19. Melby MJ, Bergman K, Ramos T, et al. Acute adrenal insufficiency mimicking septic shock: a case report. Pharmacotherapy 1988; 8:69–71. 20. Gabbay DS, Cunha BA. Pseudosepsis secondary to bilateral adrenal hemorrhage. Heart Lung 1998; 27:348–351. 21. Wilson PG, Manji M, Neoptolemos JP. Acute pancreatitis as a model of sepsis. J Antimicrob Chemother 1998; 41(suppl A):51–63. 22. Cunha BA. Fever in the critical care unit. Crit Care Clin 1998; 14:1–14. 23. Opal SM, Cohen J. Clinical gram-positive sepsis: does it fundamentally differ from gramnegative bacterial sepsis. Crit Care Med 1999; 27:1608–1616. 24. Levy B, Bollaert PE. Clinical manifestations and complications of septic shock. In: Dhainaut JG, Thijs LG, Park G, eds. Septic Shock. Philadelphia: WB Saunders, 2000: 339–352. 25. Court O, Kumar A, Parrillo JE, et al. Clinical review: myocardial depression in sepsis and septic shock. Crit Care 2002; 6:500–508. 26. Ristuccia PA, Hoeffner RA, Digamon-Beltran M, et al. Detection of bacteremia by buffy coat smears. Scand J Infect Dis 1987; 19:215–217. 27. Llewelyn M, Cohen J. Intern Sepsis Forum. Diagnosis of infection in sepsis. Intensive Care Med 2001; 27(suppl 1):S10–S32. 28. Mieszczanska H, Lazar J, Cunha BA. Cardiovascular manifestations of sepsis. Infect Dis Prac 2003; 27:183–186. 29. Takala A, Nupponen I, Kylanpaa-Back ML, et al. Markers of inflammation in sepsis. Ann Med 2002; 34:614–623. 30. Povoa P. C-reactive protein: a valuable marker of sepsis. Intensive Care Med 2002; 28:235–243. 31. Cunha BA. Antibiotic treatment of sepsis. Med Clin North Am 1995; 79:551–558. 32. Cunha BA. Antibiotic Essentials. (5th Ed) Royal Oak Physicians Press, 2006. 33. Kollef MH. Inadequate antimicrobial treatment: an important determinant of outcome for hospitalized patients. Clin Infect Dis 2000; 31(suppl 4):S131–S138. 34. Finch RG. Empirical choice of antibiotic therapy in sepsis. JR Coll Physicians Lond 2000; 34:528–532. 35. Bochud PY, Glauser MP, Calandra T, et al. Antibiotics in sepsis. Intensive Care Med 2001; 27(suppl 1):S33–S48. 36. Lepper PM, Held TK, Schneider EM, et al. Clinical implications of antibiotic-induced endotoxin release in septic shock. Intensive Care Med 2002; 28:824–833. 37. Danner RL, Elin RJ, Hosseini JM, et al. Endotoxemia in human septic shock. Chest 1991; 99:169–175. 38. Dellinger RP. Current therapy for sepsis. Infect Dis Clin North Am 1999; 13:495–509. 39. Vincent JL. International Sepsis Forum. Hemodynamic support in septic shock. Intensive Care Med 2001; 27(suppl 1):S80–S92. 40. Carlet J. Antibiotic management of severe infections in critically ill patients. In: Dhainaut JG, Thijs LG, Park G, eds. Septic Shock. Philadelphia: WB Saunders, 2000:445–460. 41. Marshall JC. Control of the source of sepsis. In: Vincent JL, Carlet J, Opal SM, eds. The Sepsis Text. Boston: Kluwer Academic Publishers, 2002:525–538. 42. Callister ME, Evans TW. Haemodynamic and ventilatory support in severe sepsis. JR Coll Physicians Lond 2000; 34:522–528. 43. Annane D. Corticosteroids for septic shock. Crit Care Med 2001; 29(suppl):S117–S120. 44. Healy DP. New and emerging therapies for sepsis. Ann Pharmacother 2002; 36:648–654.

PART II: CLINICAL SYNDROMES

5 Meningitis and Its Mimics in the Critical Care Unit Burke A. Cunha Infectious Disease Division, Winthrop-University Hospital, Mineola, and State University of New York School of Medicine, Stony Brook, New York, U.S.A.

OVERVIEW There are several diagnostic difficulties in patients presenting with the possibility of acute bacterial meningitis. Critically ill patients with meningitis are usually transferred to the critical care unit (CCU) for intensive supportive care. Meningitis may be mimicked by a variety of infectious and noninfectious disorders. The mimics of meningitis are readily ruled out on the basis of the history/physical exam and, if any doubt remains, then a lumbar puncture with cerebrospinal fluid (CSF) analysis will include or exclude the diagnosis of acute bacterial meningitis. Early and appropriate empiric antimicrobial therapy of acute bacterial meningitis in the CCU may be lifesaving. In contrast to the differential diagnostic problem of encephalitis in the CCU, acute bacterial meningitis in the CCU is not usually a diagnostic problem but is primarily a therapeutic problem. Acute bacterial meningitis is, in the main, caused by bacterial neuropathogens. Acute bacterial meningitis occurs in normal and compromised hosts and may be acquired naturally or as a complication of open head trauma or neurosurgical procedures. Regardless of the pathogen or mode of acquisition, the definitive diagnosis of acute bacterial meningitis rests on analysis of the CSF profile and gram stain/ culture of the CSF. Acute bacterial meningitis in normal and compromised hosts presents clinically with meningeal irritation, i.e., nuchal rigidity. Nuchal rigidity must be differentiated from other causes of neck stiffness, i.e., meningismus associated with the mimics of meningitis. There are relatively few nonbacterial causes of meningitis, and it is important to differentiate aseptic or viral meningitis from bacterial meningitis. In general, patients with aseptic or viral meningitis are less critically ill than are those with acute bacterial meningitis. Patients ill enough to be admitted to the CCU usually are more likely to have bacterial versus viral meningitis. Aseptic viral meningitis may be diagnosed by analysis of the CSF profile, as well as specific viral culture/polymerase chain reaction (PCR) determinations. Patients with acute meningitis, either bacterial or viral, will have various degrees of nuchal rigidity with intact mental status. Patients with mental confusion, i.e., encephalopathy, have 81

82

Cunha

encephalitis, and these patients do not have nuchal rigidity. Central nervous system (CNS) infection caused by a few organisms, i.e., Herpes simplex virus (HSV)-1, Mycoplasma pneumoniae, and Listeria monocytogenes, may present with a combination of stiff neck and mental confusion, i.e., meningoencephalitis. Any patient with fever and otherwise unexplained neck stiffness should have a lumbar puncture performed to confirm the diagnosis of acute bacterial meningitis. If acute bacterial meningitis is suspected, lumbar puncture should be performed prior to head computed tomography (CT)/magnetic resonance imaging (MRI) (1–6). Therefore, the challenge of meningitis in the CCU setting is to arrive at a correct diagnosis by ruling out the noninfectious mimics of meningitis, and then differentiating viral meningitis from bacterial meningitis. Patients with signs of meningeal irritation and mental confusion, i.e., meningoencephalitis, are diagnosed on the basis of the CSF profile and extra-CNS signs, symptoms, and/or laboratory abnormalities. The objective of arriving at a presumptive diagnosis of acute bacterial meningitis is to begin appropriate empiric therapy as soon as possible. Appropriate empiric therapy for acute bacterial meningitis is determined by predicting the likely range of pathogens. In acute bacterial meningitis, the most likely pathogen is determined by the age of the patient, mode of onset, epidemiological history/predisposing factors, physical signs, e.g., rash, rhinorrhea, cranial nerve abnormalities, etc., specific host defense defects and associated underlying disorders, and the morphology/arrangement of organisms seen on the gram stain of the CSF (1,2,4).

CLINICAL DIAGNOSIS OF ACUTE BACTERIAL MENINGITIS Overview Excluding open CNS trauma or neurosurgical procedures, bacteria causing acute meningitis reach the CSF hematogenously. Many bacteria have a bacteremic potential, i.e., bacteremias are part of their infection process, but relatively few are able to cross the blood–brain barrier and cause meningitis. Acute bacterial meningitis usually involves the leptomeninges or the covering of the brain. Leptomeningeal irritation is responsible for the nuchal rigidity, Kernig’s and Brudzinski’s signs associated with acute bacterial meningitis (7,8). Because the leptomeninges cover the brain parenchyma, meningitis is not associated with changes in mental status that require parenchymal invasion. The majority of pathogens causing acute bacterial meningitis are respiratory tract organisms. Acute bacterial meningitis may also result from contiguous spread from a local source in close proximity to the brain. Infections that cause meningitis by contiguous spread include sinusitis or mastoiditis. Cracks in the cribriform plate are another example of a mode of entry via a contiguous bacterial source. Meningitis may also occur by hematogenous spread of nonrespiratory pathogens, e.g., Listeria, Escherichia coli, and Staphylococcus aureus, as part of secondary bacteremia with CNS seeding. Acute bacterial endocarditis due to S. aureus is not infrequently complicated by acute purulent bacterial meningitis as a suppurative complication (9). The insertion of CNS shunts for hydrocephalus/ increased intracranial pressure, if complicated by meningitis, reflects either the flora of the skin introduced during the insertion process, or the flora at the distal end of the shunt, i.e., a ventricular peritoneal shunt. Open head trauma introduces the bacteria into the CSF/brain parenchyma (Table 1) (1–5,10–13). Meningoencephalitis due to L. monocytogenes is recognizable by clues from the CSF profile and is common in the elderly/immunosuppressed. M. pneumoniae

Meningitis and Its Mimics in the Critical Care Unit

83

Table 1 Symptoms and Signs of Acute Bacterial Meningitis Symptoms

Signs

Headache Photophobia Nausea and vomiting

Fever Meningismus Kernig’s sign Brudzinski’s sign Acute deafness Cranial nerve palsies Seizures

meningoencephalitis is being recognized as part of the clinical presentation of M. pneumoniae atypical pneumonia. M. pneumoniae meningoencephalitis occurs in patients with Mycoplasma community–acquired pneumonia with very high cold agglutinin levels ( >1:512). The viruses, e.g., enteroviruses, that cause meningitis are relatively few compared to their bacterial counterparts. Some viruses, i.e., HSV-1 cause a spectrum of CNS infections in normal hosts from aseptic meningitis to encephalitis. Partially treated meningitis is bacterial meningitis following initial treatment for meningitis. Partially treated bacterial meningitis is diagnosed by history, and findings in the CSF, i.e., pleocytosis with a variably decreased glucose and a moderately elevated CSF lactic acid (4–6 mmol/L). Partially treated meningitis requires retreatment with antimicrobials with the same spectrum and dosage as to treat acute bacterial meningitis (1–3,14,15).

THE MIMICS OF MENINGITIS Overview Because a stiff neck or nuchal rigidity is the hallmark of acute bacterial meningitis, any condition that is associated with neck stiffness may mimic meningitis. Patients with acute torticollis, muscle spasm of the head/neck, cervical arthritis, or meningismus due to a variety of head and neck disorders can all mimic bacterial meningitis. Fortunately, most of these causes of neck stiffness or meningismus are not associated with fever. Fever plus nuchal rigidity is the distinguishing hallmark of acute bacterial meningitis. It may be difficult in elderly patients to rule out meningitis on the basis of fever and nuchal rigidity alone because many elderly individuals have fever due to a variety of non-CNS infections, and may have a stiff neck due to cervical arthritis. In such situations, analysis of the CSF profile will readily distinguish the mimics of meningitis from actual infection (1–5). Noninfectious Mimics of Acute Bacterial Meningitis Disorders that commonly may be mistaken for meningitis include drug-induced meningitis, meningeal carcinomatosis, serum sickness, collagen vascular diseases, granulomatous angiitis of the CNS, Behc¸et’s disease, systemic lupus erythematosus (SLE), and neurosarcoidosis. The diagnostic approach to the mimics of meningitis is related to the clinical context in which they occur. For example, lupus cerebritis would rarely present as the sole manifestation of SLE. Similarly, with Behc¸et’s disease, patients developing neuro-Behc¸et’s disease have established Behc¸et’s, and have multiple manifestations, which should lead the clinician to suspect the diagnosis in

84

Cunha

such a patient. Similarly, with neurosarcoid, the presentation is usually subacute or chronic rather than acute, and occurs in patients with a known history of sarcoidosis (Table 2) (16–21). Drug-Induced Meningitis Drug-induced meningitis may present with a stiff neck and fever. The time of meningeal symptoms after consumption of the medication is highly variable. The most Table 2 Mimics of Acute Bacterial Meningitis Drug-induced aseptic meningitis Toxic/metabolic abnormalities NSAIDs OKT13 ATG TMP-SMX Azathioprine CNS vasculitis SLE cerebritis Sarcoid meningitis Bland emboli from SBE or marantic endocarditis (nonbacterial thrombocytic endocarditis) Tumor emboli Primary or metastatic CNS malignancies (meningeal carcinomatosis) AML ALL Hodgkin’s lymphoma Non-Hodgkin’s lymphoma Melanoma Breast carcinomas Bronchogenic carcinomas Hypernephromas (renal cell carcinomas) Germ cell tumors Legionnaires’ disease Posteria fossa syndrome Subarachnoid hemorrhage Intracerebral hemorrhage CNS leukostasis Thrombocytopenia DIC Abnormal platelet function Coagulopathy CNS metastases Embolic and thrombotic strokes Partially treated bacterial meningitis Meningoencephalitis Abbreviations: AML, acute myelogenous leukemia; ALL, acute lymphoblastic leukemia; DIC, disseminated intravascular coagulation; CNS, central nervous system; SLE, systemic lupus erythematosus; NSAIDs, nonsteroidal inflammatory drugs; ATG, antithymoglobulin; SBE, subacute endocarditis; TMP-SMX, trimethoprim-sulfamethoxazole. Source: From Refs. 1–6.

Meningitis and Its Mimics in the Critical Care Unit

85

common drugs associated with drug-induced meningitis include use of nonsteroidal inflammatory drugs. In addition, trimethoprim alone (TMP-SMX), and to a lesser extent, azathioprine may present as a drug-induced aseptic meningitis. Leukocytosis in the CSF with a polymorphonuclear predominance is typical with drug-induced meningitis, and the clinical clue to the presence of drug-induced meningitis is the presence of eosinophils in the CSF. In drug-induced meningitis, the CSF also contains increased protein, but the CSF glucose is rarely decreased. RBCs or an increased CSF lactic acid level are not features of drug-induced meningitis. Treatment is discontinuation of the offending agent (1,16,17).

Serum Sickness Serum sickness is a systemic reaction to the injection of serum-derived antitoxin derivatives. Because such toxins are not used much anymore, serum sickness is now most commonly associated with the use of certain medications, including b-lactam antibiotics, sulfonamides, and streptomycin among the antimicrobials. Nonantimicrobials associated with serum sickness include hydralazine, alpha methyldopa, propanolol, procainamide, quinidine, phenylbutazone, naproxen, catapril, and diphenyl hydantoin. Symptoms typically begin about two weeks after the initiation of drug therapy, and are characterized by fever, arthralgias/arthritis, and immune complex–mediated renal insufficiency. Urticaria, abdominal pain, or lymphadenopathy may or may not be present. Neurologic abnormalities are part of the systemic picture and include a mild meningoencephalitis, which occurs early in the first few days with serum sickness. Ten percent of patients may have papilledema, seizures, circulatory ataxia, transverse myelitis, or cranial nerve palsies. The clues to serum sickness systemically are an increased sedimentation rate, a decreased serum complement, microscopic hematuria/RBC casts, and hypergammaglobulinemia. The CSF typically shows a mild lymphocytic pleocytosis, and protein is usually normal but may be slightly elevated as is the CSF glucose. The cause of the patient’s fever and meningeal symptoms may be related to serum sickness if the clinician appreciates the association of the CNS findings and extra-CNS manifestations of serum sickness. Treatment is with corticosteroids (1–4). Collagen Vascular Diseases SLE often presents with CNS manifestations ranging from meningitis to cerebritis, and encephalitis. The most frequent CNS manifestation of SLE is aseptic meningitis, which needs to be differentiated from acute bacterial meningitis. CNS manifestations of SLE usually occur in patients who have established multisystem manifestations of SLE. CNS SLE is usually present as part of a flare of SLE. SLE flare may be manifested by fever, an increase in the signs/symptoms of SLE manifested in previous flares. Laboratory tests suggesting flare include new or more severe leukopenia, thrombocytopenia, increased erythrocyte sedimentation rate (ESR), polyclonal gammopathy, and proteinuria/microscopic hematuria. The CSF in patients with SLE includes a lymphocytic predominance (usually 90%) Glucose may be #/normal " Lactic acid ~ RBCs in CSF

Meningeal carcinomatosis

History: leukemias, lymphomas, carcinomas, or without known primary neoplasm Onset: subacute/afebrile Mental status changes: þ/ Nuchal rigidity: þ/ 80% have cranial nerve involvement, (CNs III, IV, VI, VII, or VIII most common) CSF: Gram stain:  RBCs:  Protein: highly " Lactic acid: variably " Cytology: abnormal in 90%

Amebic meningoencephalitis (Naegleria fowleri)

History: recent swimming in fresh water Onset: rapid Olfactory/gustatory abnormalities: early Head MRI/CT: mass lesions CSF: RBCs: þ Glucose: # Lactic acid: variably " Gram stain: ‘‘motile WBCs’’ (ameba) on wet prep (Continued )

88 Table 3

Cunha Mimics of Acute Bacterial Meningitis (Continued )

Meningeal mimic

Differential features and diagnostic clues

Brain abscess (with ventricular leak)

Source usually suppurative lung disease (bronchiectasis), cyanotic heart disease (R ! L shunts), mastoiditis, dental abscess, etc. Head MRI/CT: mass lesions CSF: mimics bacterial meningitis (with ventricular leak) Protein: highly " Without leak: usually 10normal) Gross hematuria, pyuria, and hemoglobinuria

Abbreviations: CN, cranial nerve; CNS, central nervous system; CSF, cerebrospinal fluid; CVA, costovertebral angle; GI, gastrointestinal; HEENT, head, eyes, ears, nose and throat; RBC, red blood cell; SBE, subacute bacterial endocarditis; WBC, white blood cell; SGOT, serum glutamate oxaloacetate transaminase; SGPT, serum glutamate pyruvate transaminase.

diagnosis is made by demonstrating elevated IgM titer with or without an elevated IgG M. pneumoniae enzyme linked immunosorbent assay (ELISA) titer enzyme linked immunosorbent assay (5,12). Legionnaires’ Disease Legionnaires’ disease is an infection caused by any species of Legionella and is a systemic infection that predominantly affects the lungs. Legionnaires’ disease, as with other atypical causes of CAP, is characterized by its particular pattern of extra-pulmonary organ involvement. In patients with Legionella CAP, CNS manifestations are frequent and include varying levels of consciousness, headache, and, most commonly, encephalopathy manifested my mental confusion. The characteristic pattern of extra-pulmonary organ involvement with Legionella, in addition to CNS findings,

Encephalitis and Its Mimics in the Critical Care Unit

109

includes heart, liver, renal, and gastrointestinal involvement (Table 2). Legionella infections do not have upper respiratory tract or skin manifestations. The main cardiac manifestation of Legionella is relative bradycardia, i.e., a pulse temperature deficit (excluding patients on beta-blocker therapy, with arrhythmias, or pacemaker-induced rhythms). Gastrointestinal involvement is characterized by watery diarrhea as with M. pneumoniae. Hepatic involvement is manifested by an early mild/transient increase in the serum transaminases. In patients with atypical CAPs a slightly increased SGOT/SGPT limits diagnostic possibilities to Legionnaires’ disease, Q fever, or psittacosis. CAP due to Q fever or psittacosis present with severe headache as the primary CNS manifestation but rarely, if ever, with encephalopathy. A low serum sodium secondary to syndrome of inappropriate ADH (SIADH) is common in patients with Legionella as well as any patient with a pulmonary/CNS process, although frequently Legionella is nonspecific and diagnostically unhelpful. Other laboratory tests pointing to Legionella are an elevated creatine phosphokinase (CPK) or creactive protein (CRP) of more than 30. Urinalysis may show otherwise unexplained microscopic hematuria. Increased cold agglutinins are not a feature of Legionnaires’ disease and, if present, argue strongly against the diagnosis. Because coinfections are exceedingly rare, high cold agglutinins 1:64 in a patient with atypical CAP should point to M. pneumoniae and not a co-infection with Mycoplasma and Legionella. Relative bradycardia is also not a feature of M. pneumoniae, which is a constant finding in those with Legionnaires’ disease with or without CNS manifestations. The clinical syndromic diagnosis of Legionella is based on finding one or more of the extra-pulmonary clinical or laboratory findings mentioned and, if present, should prompt specific diagnostic testing. Legionella may be diagnosed early before treatment by direct fluorescent antibody (DFA) or respiratory secretions. Serologically, Legionella may be demonstrated by an increased IgM titer or a fourfold or greater increase between acute and convalescent IgG titers. Legionella pneumophila (serotype 01) may be diagnosed by the urinary antigen test. Legionella antigenuria is often negative early in the infectious process when the patient is in the CCU and is only positive with L. pneumophila serogroup 01. Legionella antigenuria becomes positive over time, persists for weeks after the infection, and is of more use later in the course of infection/convalescence when a retrospective diagnosis is desired. In the critical care setting, a negative initial Legionella titer or a negative Legionella antigen determination does not rule out Legionella. The presumptive diagnosis of Legionella is based on the clinical syndrome suggested by the characteristic pattern of organ involvement and laboratory tests described (Table 3) (5,14).

NONINFECTIOUS MIMICS OF ACUTE VIRAL ENCEPHALITIS Toxic/Metabolic Encephalopathy The most common cause of noninfectious acute encephalopathy in the CCU are related to acute metabolic/toxic encephalopathy usually due to one or more medications. Patients with toxic/metabolic abnormalities have no findings to suggest that an infectious etiology and toxic/metabolic encephalopathy is largely a diagnosis of exclusion. Computed tomography (CT)/magnetic resonance imaging (MRI) scans of the head are negative, and the EEG is unremarkable/nonspecific. CSF findings in toxic/metabolic encephalopathy usually contain T-lymphocyte function which may predispose them to pneumonia or bacterial meningitis superimposed on their on hepatic encephalopathy. Lumbar puncture should be obtained, if possible, using plating/clotting factors infusions prelumbar puncture because such patients often have coagulopathy related to their hepatic dysfunction. Lumbar puncture with analysis of the CSF gram stain/cultures is the only way to definitely rule out a coexisting bacterial process that could easily be masked by the superimposed hepatic encephalopathy (Table 4) (1–3). CNS Hemorrhage CVA Patients in the CCU due to a massive cerebral vascular accident (CVA) or a massive intracranial hemorrhage may be suspected on the basis of neurologic findings. CNS hemorrhage or CVA may be confirmed by head CT or MRI scans. Lumbar puncture will reveal a bloody CSF in patients with CNS hemorrhage that communicates with the ventricles. CSF gram stain and culture will be negative, but in patients with intracranial hemorrhage, there will be a CSF pleocytosis and a variably elevated protein. The CSF glucose may be decreased in proportion to the RBCs present. RBCs in the CSF actively metabolize glucose and decrease the CSF glucose via this mechanism. The CSF lactic acid will also be elevated in direct proportion to the number of red cells present, but the negative CSF gram stains/culture will rule out an infectious etiology (1–3). MISCELLANEOUS OTHER DISORDERS SLE Cerebritis There a variety of miscellaneous disorders that may present with mental confusion. Most of these are collagen vascular diseases, e.g., systemic lupus erythematosus (SLE)

Encephalitis and Its Mimics in the Critical Care Unit

111

Table 4 Noninfectious Mimics of Acute Encephalitis Disorder Primary CNS lymphomas

Metastatic lymphoma to CNS

Metastases to CNS

Squamous/small cell bronchogenic carcinoma

CSF findings þ Cytology " Protein  RBCs # Glucose Normal/" lactic acid (RBCs) þ Cytology " Protein  RBCs # Glucose N/" lactic acid (RBCs) þ Cytology

" Protein  RBCs

# Glucose N/" lactic acid (RBCs)

Breast carcinoma

Other carcinomas

Massive intracerebral hemorrhage SLE cerebritis

Granulomatous CNS angitis

Toxic/metabolic

þ Cytology " Protein  RBCs # Glucose N/" lactic acid (RBCs) þ Cytology " Protein  RBCs # Glucose N/" lactic acid (RBCs) """ RBCs/bloody tap # Glucose " Lactic acid (RBCs) Lymphocytic pleocytosis  # Glucose  # Lactic acid # CSF, C4 No pleocytosis

N glucose N lactic acid No pleocytosis N glucose No RBCs N lactic acid

Other findings Mass lesion on head CT/MRI þ HIV serology

Primary lymphoma

CXR central mass lesion ( cavitation with squamous cell carcinoma) " Caþþ (with squamous cell carcinoma) Clubbing/Hypertrophic Pulmonary Osteoarthropathy (HPO) SVC/IVC syndrome þ Bone marrow cytology with (small/squamous cell carcinomas) History/presence of breast cancer

History/presence of extra CNS primary malignancy

History/features of SLE Evidence of SLE flare (# WBC count, # C3, " ferritin) þ Head CT Magnetic Resonance Angiograhm (MRA) þ Brain biopsy Multiple medications (opiates, narcotics, sedatives, etc.)

(Continued )

112 Table 4

Cunha Noninfectious Mimics of Acute Encephalitis (Continued )

Disorder Hepatic encephalopathy

CSF findings No pleocytosis N glucose No RBCs N lactic acid

Other findings History/presence of severe/ advanced liver disease " Ammonia levels

Abbreviations: CNS, central nervous system; CSF, cerebrospinal fluid; HIV, human immunodeficiency virus; RBC, red blood cell; SLE, systemic lupus erythematosus; WBC, white blood cell; CXR, chest X-ray; HPO, hypertrophic pulmonary osteoarthropathy. Source: From Refs. 1–4.

or various causes of cerebral vasculitis. Among the collagen vascular diseases with CNS manifestations, SLE is the most common. Behc¸et’s disease, sarcoidosis, antiphospholipid syndrome, and Sjo¨gren’s occasionally mimic meningitis but not encephalitis. Lupus cerebritis among the collagen vascular diseases is the most likely to mimic encephalitis. The possibility of lupus cerebritis is suggested by the patient’s history, i.e., a long-standing history of lupus with multisystem manifestations. Lupus cerebritis may be diagnosed from head CT/MRI appearance showing bilateral abnormalities over the surface of both hemispheres. The diagnosis of lupus cerebritis may be confirmed by demonstrating decreased C4 level in the CSF. Lupus cerebritis usually occurs as part of a SLE flare. SLE flare is suggested by the presence of leukopenia, decreased complement levels, or an increased ferritin level. SLE may also present with psychosis, seizure, or CVA (5,15,16).

CNS Vasculitis Among the vasculitides that have CNS manifestations are periarteritis nodosa, Churg–Strauss granulomatosis, CNS angitis, and temporal arteritis. Patients with these disorders usually present with headache mimicking meningitis rather than encephalitis. Of this group, granulomatous CNS angiitis involving the leptomeninges may present with encephalitis. Diagnosis is by CT angiography or brain biopsy. The erythrocyte sedimentation rate (ESR) is highly elevated, and the serological tests for other collagen vascular diseases, e.g., SLE are negative. Neurologic involvement with sarcoidosis is infrequent but occurs in 5% to 10% of cases. The most common manifestation of neurosarcoidosis is chronic basilar meningitis. Cranial neuropathy, most frequently central seventh nerve palsy, with optic nerve involvement is the characteristic finding in neurosarcoidosis. Neurosarcoidosis may also present as AIDS or septic meningitis but encephalitis is rarely, if ever, a feature of neurosarcoidosis (1,15,17,18).

CNS Malignancies Primary or metastatic disease to the CNS often presents as mental confusion. Primary CNS lymphoma occurs in patients with severely impaired T-cell function, e.g., human immunodeficiency virus (HIV). It is difficult in such patients without imaging to differentiate the encephalopathy due to HIV from HIV with superimposed CNS lymphoma. With HIV encephalopathy, the head CT/MRI shows no

Encephalitis and Its Mimics in the Critical Care Unit

113

mass lesions, but with HIV and CNS lymphoma, there are mass lesions seen with neuroimaging studies. A variety of neoplasms metastasizes to the CNS and may present with encephalopathy. Common among these are the bronchogenic carcinomas, particularly squamous cell carcinoma and small cell carcinoma. Breast carcinoma is a frequent cause of CNS metastases. Patients with a history of breast cancer and encephalopathy should be viewed as potentially having encephalopathy due to CNS metastases. Metastatic lymphomas are also a common cause of metastatic disease to the CNS that present with mental confusion. Although virtually any tumor may metastasize to the CNS, other malignancies do so only rarely. The suspicion of CNS metastases based on a tumor is based on history, and the diagnosis may be confirmed by cytology of the CSF and/or brain biopsy. West Nile encephalitis (WNE) may also occur in patients with cancer, mimicking CNS metastases (19–21).

VIRAL CAUSES OF ACUTE ENCEPHALITIS IN THE CCU The viruses responsible for acute encephalitis may be classified in several ways. Acute encephalitis may be classified as either being transmitted or not transmitted by arthropod vectors. Encephalitic viruses may also be classified according to season of peak occurrence, i.e., those having a seasonal, e.g., arboviruses, or nonseasonal occurrence, e.g., HSV-1 encephalitis. The viruses causing acute encephalitis have been named according to their original location, i.e., Powassan virus, West Nile virus (WNE), etc., of their isolation or their host, i.e., the equine encephalitides. Other neurotropic viruses causing encephalitis include the agent of Colorado tick fever, rabies, etc. The viruses causing acute encephalitis have many different physiochemical characteristics, come from different families, and are transmitted differently to humans. They also differ in their rapidity of onset, severity, and lethality. Encephalitis viruses may also be considered clinically as those that cause only encephalitis, e.g., HSV-1 and those that cause encephalitis with other extra CNS findings, e.g., WNV with encephalitis plus flaccid paralysis. In compromised hosts with impaired T-cell immunity, Epstein–Barr virus (EBV), human herpes virus-6 (HHV-6) and CMV may present with encephalitis. As there are only a limited number of antiviral agents effective against all the viral etiologies of viral encephalitis, drugs are available only for HSV and CMV. The clinical diagnostic approach in the CCU has a twofold purpose. The first clinical task is to eliminate disorders that may mimic encephalitis. The clinicians then can treat the few underlying disorders that are nonviral and for which there is effective therapy, e.g., Listeria, Legionella, and M. pneumoniae meningoencephalitis. The last task faced by the clinician is to try and differentiate the treatable causes of viral encephalitis, e.g., HSV or CMV, from the other nontreatable causes of viral encephalitis (Tables 5–8) (1–5,23–45).

CLINICAL DIAGNOSTIC APPROACH Treatable Noninfectious Mimics of Viral Encephalitis Among the noninfectious mimics of encephalitis, hepatic encephalopathy, SLE cerebritis, granulomatous CNS angiitis, some types of metastatic carcinomas to the CNS, and metastatic lymphomas to the CNS may be responsive to treatment. Other noninfectious mimics of encephalitis that are usually not responsive to therapeutic interventions are massive CVAs/intracranial hemorrhage.

114

Cunha

Table 5 Causes of Acute Viral Encephalitis HSV-1/HSV-2 VZV HIV Arboviruses US Western equine Eastern equine St. Louis California group Venezuelan equine Powassan Colorado tick fever West Nile virus Influenza A LCM Enteroviruses CMVa Mumps Measles Rubella EBV Adenoviruses Toxoplasma gondiia Rabies a Only in human immunodeficiency virus, organ transplants. Abbreviations: HSV-1/HSV-2, herpes simplex types 1 and 2; VZV, varicellazoster virus; HIV, human immunodeficiency virus; LCM, lymphocytic choriomeningitis; CMV, cytomegalovirus; EBV, Epstein–Barr virus. Source: From Refs. 1–5.

Table 6 Geographical Distribution of Acute Arboviral Encephalitis North America

South America

Eastern Europe Africa Asia

Australia

EEE WEE StLE CE LAC POW WNE VEE EEE StLE TBE WNE JE RSSE WNE MVE

Abbreviations: EEE, eastern equine encephalitis; WEE, western equine encephalitis; StLE, St. Louis encephalitis; CE, California encephalitis; LAC, La Crosse; POW, Powassan; WNE, West Nile encephalitis; VEE, Venezuelan equine encephalitis; TBE, European tick-borne encephalitis; JE, Japanese B encephalitis; RSSE, Russian spring-summer encephalitis; MVE, Murray Valley encephalitis. Source: From Refs. 1–3, 54.

Age group affected

Children 10 yrs

Children and adults (70% mortality)

Children and adults

Children and adults

Adults (60 yrs)

Children (30) (Table 4) (6,18). Disorders Associated with Decreased PMN Function Most disorders associated with neutropenia are the result of chemotherapy. When the peripheral PMN count drops below 1 K/mm3, the probability of infection increases greatly. Patients with neutropenia are predisposed to Pseudomonas aeruginosa bacteremia or Candidemia. Patients with prolonged neutropenia also are predisposed to Aspergillus and other fungal infections. Patients with neutropenia do not ordinarily present with pneumonia even though their PMN counts are very low, but present with bacteremia or fungemia (6,19). Severe CAP Accompanied by Cavitation The presence of cavitation on the chest X-ray or computed tomography chest scan is an important diagnostic clue to the etiology of the pulmonic infectious process. Pneumonias with cavitation tend to be severe because they represent a necrotic invasive pulmonary process. Severe CAP that presents with cavitation may be

Severe Community-Acquired Pneumonia in the Critical Care Unit

163

Table 4 Diagnostic Approach to Bilateral Diffuse Pulmonary Infiltrates Without oxygen diffusion defect (N/# pO2/N A-a gradient) Aspiration pneumonia Pulmonary edema/LHF Pulmonary hemorrhage Typical bacterial pneumonitis Atypical bacterial pneumonitis Advanced pulmonary tuberculosis Fungal pneumonias Rickettsial pneumonias Parasitic pneumoniasa Radiation pneumonitis Pulmonary drug reaction Noncardiogenic pulmonary edema Leukostasis

With profound oxygen diffusion defect (### pO2/" A-a gradient >30) PCP CMV HSV-1 HHV-6 Influenza RSV ARDS BOOP

a

Excluding PCP. Abbreviations: ARDS, acute respiratory distress syndrome; BOOP, bronchiolitis obliterans– organizing pneumonia; CMV, cytomegalovirus; HSV, herpes simplex virus; PCP, Pneumocystis (carinii) jiroveci pneumonia; LHF, left heart failure; HHV, human herpes virus; RSV, respiratory syncytial virus. Source: From Ref. 6.

approached clinically by the rapidity of the cavitary process. Cavitation within 72 hours in a patient with pneumonia is limited to S. aureus or S. aeruginosa pneumonias. Patients with S. aureus pneumonias are almost always those that follow viral influenza pneumonia. The presentation of influenza pneumonia is that of a viral pneumonia or may be accompanied simultaneously by pulmonary infiltrates usually due to S. aureus. The third clinical presentation of viral pneumonia is for an initial viral influenza presentation followed by a period of improvement followed subsequently by superimposed focal or segmental pneumonia usually due to S. pneumoniae or H. influenzae. Therefore, in a patient who presents with viral influenza pneumonia during the influenza season, the chest X-ray is unremarkable even though oxygen diffusion defects are present but pulmonary infiltrates are virtually absent in the early phases. If such a patient has pulmonary infiltrates, then a superimposed bacterial pneumonia is the working diagnosis. Because S. pneumoniae and H. influenzae usually follow a period of improvement and are not accompanied by cavitation, the diagnostic possibilities of an influenza presentation with rapid cavitation within 24 hours limit the diagnosis to superimposed S. aureus pneumonia. Pseudomonas pneumonia is rare and nearly always fatal, and it occurs only in the setting of chronic bronchiectasis or cystic fibrosis. Normal hosts do not present with P. aeruginosa CAP. Patients presenting with bilateral interstitial infiltrates and a diffusion defect may have a noninfectious disorder, i.e., BOOP or ARDS. If the etiology of the infiltrate is infectious, the differential diagnosis is limited to PCP or viral causes. One of the common clinical scenarios in a nonorgan transplant patient presenting with CAP is a HIV patient with PCP/CMV. In transplant patients or those with impaired CI, CMV pneumonia may present as an isolated clinical entity as severe CAP. However, in the HIV patient, PCP is the predominant cause of the patient’s diffusion defect and CMV is a secondary or suppressed pathogen, which may be demonstrated as an incidental finding in three quarters of the patients

164

Cunha

with PCP and HIV. The bystander role of CMV is indicated by the fact that if the PCP is treated in an HIV patient, the patient gets well and the CMV does not require separate treatment as it does in a transplant patient with CMV pneumonia (1,17). Cavitation occurring after three to five days points to K. pneumoniae as the pathogen. K. pneumoniae occurs almost exclusively in patients with chronic alcoholism. Therefore, the clinical history plus the speed of cavitation will point to the diagnosis, which is easily confirmed by gram stain/culture of the sputum and/or blood cultures. The K. pneumoniae CAP typically presents as severe CAP. Cavitation after five to seven days is associated with aspiration pneumonia. Such patients will not present as severe CAP unless the aspiration is bilateral and massive, or if the aspiration is more limited but superimposed upon limited pulmonary function. These patients usually present with pneumonia that becomes more severe as cavitation becomes apparent after the first week of hospitalization. In massive bilateral aspiration, severe CAP is the usual clinical presentation (19,20). Empiric Therapy for CAP Appropriate empiric therapy depends upon identifying the most likely pathogen in the patient presenting with severe CAP in the CCU. The pathogen is determined primarily by host factors. The presentation of severity may be manifested by focal and segmental infiltrates unaccompanied by diffusion defect or bilateral interstitial infiltrates with or without accompanying oxygen diffusion defect. The patient’s history is important in identifying previously diagnosed disorders that are associated with specific immune defects that, combined with the X ray presentation and the presence or absence of profound hypoxemia, helps limit differential diagnosis possibilities (10,21–23). A patient presenting with severe CAP, who is apparently a normal host with focal or segmental infiltrates, should be treated for the usual typical and atypical pathogens causing CAP. Therapy should be started as soon as the diagnosis of CAP is made (20,24). Normal hosts presenting with near normal chest X-ray and profound hypoxemia should be considered as having viral influenza or PCP. If the severe pneumonia occurs during influenza season, then influenza is a likely, diagnostic possibility. If the CAP occurs during spring, summer, or fall, then PCP is likely, particularly if accompanied by isolated cytopenias. PCP is becoming more common as more people in the general population develop HIV. PCP is an HIV-defining illness and is not an infrequent pulmonary presentation for HIV. The influenza patient is treated with antivirals, and antimicrobial therapy is not needed unless there is simultaneous infection with S. aureus or subsequent infection due to S. pneumoniae or H. influenzae. PCP should be treated initially with TMP-SMX or pentamidine accompanied by steroids as the diagnostic work up is in progress. Patients who are on steroids or immunosuppressive therapy, as well as organ transplants, may present with focal or segmental pneumonias that are not accompanied by diffusion defects. Bacterial pathogens should be covered empirically in these patients, even though the diagnostic work up proceeds to exclude such causes as Aspergillus. Because the number of fungal pathogens is extensive, a tissue biopsy is needed upon which to base specific antifungal therapy. Immunosuppressed patients/organ transplants presenting with bilateral symmetrical interstitial infiltrates may be approached in two categories, i.e., those with an oxygen diffusion defect and those without an oxygen diffusion defect. In such patients, the absence of a diffusion defect suggests pulmonary hemorrhage,

Subacute/chronic

Bilateral diffuse

Oseltamivir þ either amantadine or rimantadine INH, rifampin, ethambutol, PZA TMP-SMX or minocycline Caspofungin or voriconazole HSV (acyclovir), CMV (ganciclovir or valgansadovir), RSV (ribaririn). Influenza A Tuberculosis Nocardia Aspergillus HSV-1, CMV, RSV, HHV-6

Lung biopsy Lung biopsy Sputum DFA or serology or cytology on bronchoscopy

þ

CMV

Sputum þ for AFB

Cysts in trans bronchial biopsy BAL þ for CMV cytology Sputum DFA

þ

‘‘Respiratory quinolone’’ or b-lactamþdoxycycline ‘‘Respiratory quinolone’’ or b-lactamþdoxycycline TMP-SMX atoragone, or pentamidine Ganciclovir or valgangciclovir

Empiric therapy

Typical/atypical pneumonia Typical/atypical pneumonia PCP

Likely diagnosis



Sputum/blood cultures

Sputum/blood cultures

Diagnostic test





Oxygen diffusion defect: # pO2/ " A-a gradient (>30)

Abbreviations: HI, humoral immunity; CMI, cellular mediated immunity; CMV, cytomegalovirus; HSV, herpes simplex virus; AFB, acid-fast bacilli; PCP, BAL, bronchoalveolar lavage; DFA, direct fluorescent antibody; RSV, respiratory syncytial virus; HHV, human herpes virus; PZA, pyrazinamide; TMP-SMX, trimethoprim-sulfametoxozole. Source: From Refs. 1, 6, 20.

Subacute/chronic

Focal/segmental

Acute

Bilateral diffuse

# HI/CMI

Acute

Focal/segmental

# CMI

Acute

Focal/segmental

None/ # HI

Onset

Infiltrate

Host defense defect

Table 5 Clinical Approach to Community-Acquired Pneumoniae in Compromised Hosts

Severe Community-Acquired Pneumonia in the Critical Care Unit 165

166

Cunha

pulmonary embolus, or another noninfectious process. In this patient group, with bilateral infiltrates accompanied by a profound oxygen diffusion defect, viral pneumonias and PCP are the most likely diagnostic infectious possibilities. Noninfectious causes of oxygen defects and bilateral pulmonary infiltrates include BOOP and ARDS, which should be relatively straightforward diagnoses (Table 5) (6,17,20). In conclusion, the clinician should not use a ‘‘shot gun’’ approach to treating patients with severe CAP, because of the mistaken notion that there are many diagnostic possibilities. The diagnostic process is concerned with limiting diagnostic possibilities to one or two possible etiologies based on a syndromic analysis of the history, physical, and laboratory abnormalities, as well as the chest X-ray appearance and the presence or absence of an oxygen diffusion defect. Except in patients with decreased CMI, focal or segmental defects may be treated the same as in normal hosts with antibiotics that are active against typical and atypical pathogens. No unusual organisms will be missed using this approach. Patients with decreased CMI are more complex and may also present with the pathogens that affect the normal population. If patients with decreased CMI present with an acute, severe CAP, then the same therapeutic approach is used as in normal patients, i.e., use an antibiotic, e.g., respiratory quinolone or combination therapy that is effective against both typical and atypical pathogens. Usually the focal and segmental infiltrates in patients with decreased CMI are subacute or chronic, and they do not usually become diagnostic problems in the setting of severe CAP. The patients developing focal infiltrates due to Aspergillus do so over weeks rather than days. Clinicians should be aware of the noninfectious mimics of pneumonia both in the normal and compromised hosts. The mimics of pneumonia can usually be excluded by history and routine laboratory tests. Transbronchial or open lung biopsy is necessary when analysis fails or the patient is not responding to appropriate antimicrobial therapy. Compromised hosts respond more slowly than normal hosts to appropriate therapy. Normal hosts with severe CAP may show improvement in three to five days, but compromised hosts may take a week or more to show improvement. The duration of therapy in compromised hosts is necessarily longer because of their impaired host defenses. Normal hosts are usually treated for 10 to 14 days, whereas pneumonias in the compromised host are often treated for two to three weeks (20–22). The prognosis in CAP is a function of the same determinants that make CAP severe, i.e., host factors. Delay in therapy can make the prognosis worse (25–30).

REFERENCES 1. Cunha BA. Severe community-acquired pneumonia: determinants of severity and approach to therapy. Infect Med 2005; 22:53–58. 2. Moine P, Vercken J-P, Chevret S, et al. Severe community-acquired pneumonia. Chest 1994; 105:1487–1495. 3. Lim WS, van der Eerden M, et al. Defining community acquired pneumonia severity on presentation to hospital: an international derivation and validation study. Thorax 2003; 58:377–382. 4. Cunha BA. Severe community-acquired pneumonia. Crit Care Clin 1998; 14:105–118. 5. Recognizing and managing severe community-acquired pneumonia. Br J Hosp Med. 2006; 67:76–78. 6. Cunha BA. Pulmonary infections in the compromised host. Infect Dis Clin 2001; 16:591–612.

Severe Community-Acquired Pneumonia in the Critical Care Unit

167

7. van Riemsdijk, van Overbeeke IC, van den Berg B. Severe Legionnaire’s disease requiring intensive care treatment. Neth J Med 1996; 49:185–188. 8. Janssens JP, Gauthey L, Hermmann F, et al. Community-acquired pneumonia in older patients. J Am Geriatr Soc 1996; 44:539–544. 9. Dean NC, Silver MP, Bateman KA, et al. Decreased mortality after implementation of a treatment guideline for community-acquired pneumonia. Am J Med 2001; 110:451–457. 10. Cunha BA. Community-acquired pneumonias: reality revisited. Am J Med 2000; 108: 436–437. 11. Wara DW. Host defense against Streptococcus pneumoniae: the role of the spleen. Rev Infect Dis 1981; 3:299. 12. Cunha BA. Infections in acutely ill non-leukopenic compromised hosts with diabetes mellitus, SLE, asplenia, or on steroids. Crit Care Clin 1998; 8:263–282. 13. Gopal V, Bisno AL. Fulminant pneumococcal infection in ‘‘normal’’ asplenic hosts. Arch Intern Med 1977; 137:1526. 14. Cunha BA. Community-acquired pneumonias in SLE. J Crit Illness 1997; 13:779–783. 15. Jong GM, Hsiue TR, Chen CR, et al. Rapidly fatal outcome of bacteremic Klebsiella pneumonia in alcoholics. Chest 1995; 107:214–217. 16. Johnson DH, Cunha BA. Infections in alcoholic cirrhosis. Infect Dis Clin 2001; 16:363–372. 17. Cunha BA. Community-acquired pneumonia in HIV patients. Clin Infect Dis 1999; 28: 410–411. 18. Cunha BA. Severe community-acquired pneumonia. J Crit Illness 1997; 12:711–721. 19. Cunha BA. Differential diagnosis of pneumonia. In: Brandstetter R, Cunha BA, Keretsky M, eds. The Pneumonias. New Jersey: Medec Books, Oradell:1993. 20. Cunha BA, Essentials of Antimicrobial Therapy (5th Ed) Birmingham, Michigan: Physicians Press, 2006. 21. Cunha BA. Empiric therapy of community-acquired pneumonia. Chest 2004; 125:1913–1919. 22. Cunha BA. Antibiotic treatment of severe community-acquired pneumonia. Semin Respir Crit Care Med 2000; 21:61–69. 23. Cunha BA. Clinical relevance of penicillin resistant Streptococcus pneumoniae. Semin Respir Infect 2002; 17:204–214. 24. Heath CH, Grove DI, Looke DF. Delay in appropriate therapy of Legionella pneumonia associated with increased mortality. Eur J Clin Microbiol Infect Dis 1996; 5:286–290. 25. MacFarlane JT, Finch RG, Ward MJ, et al. Hospital study of adult community-acquired pneumonia. Lancet 1982; 2:255–258. 26. Klimek JJ, Ajemian E, Fontecchio S, et al. Community-acquired bacterial pneumonia requiring admission to hospital. Am J Infect Control 1983; 11:79–82. 27. Leroy O, Santre´ C, Beuscart C, et al. A five-year study of severe community-acquired pneumonia with emphasis on prognosis in patients admitted to an intensive care unit. Intensive Care Med 1995; 21:24–31. 28. Angus DC, Marrie TJ, Obrosky DS, et al. Severe community-acquired pneumonia. Am J Respir Crit Care Med 2002; 166:717–723. 29. Martinez FJ. Monotherapy versus dual therapy for community-acquired pneumonia in hospitalized patients. Clin Infect Dis 2004; 38:328–340. 30. Cunha BA. Empiric antibiotic therapy for community-acquired pneumonia: guidelines for the perplexed? Chest 2004; 125:1913–1921.

9 Nosocomial Pneumonia in the Critical Care Unit Emilio Bouza Clinical Microbiology and Infectious Diseases Department, Hospital General Universitario ‘‘Gregorio Maran˜o´n,’’ Universidad Complutense, Madrid, Spain

Almudena Burillo Department of Clinical Microbiology, Hospital Madrid-Monteprı´ncipe, Madrid, Spain

Marı´a V. Torres Clinical Microbiology and Infectious Diseases Department, Hospital General Universitario ‘‘Gregorio Maran˜o´n,’’ Universidad Complutense, Madrid, Spain

OVERVIEW Introduction Nosocomial pneumonia or hospital-acquired pneumonia (HAP) is defined as pneumonia that appears 48 hours or more after hospitalization. In this definition, it is assumed that the patient was not incubating the causative microorganism when admitted to the hospital. Patients with HAP may be managed in a ward or when the illness is severe in the intensive care unit (ICU). Ventilator-associated pneumonia (VAP) refers to pneumonia that begins and develops after endotracheal intubation (1,2). However, a patient who has just undergone tracheotomy yet is not on a ventilator is similarly susceptible to VAP. Thus, a more appropriate term would be ‘‘endotracheal-tube-associated pneumonia.’’ In this chapter we have opted, nevertheless, to use the traditional term. Epidemiology HAP is currently the second most common nosocomial infection in North America and is associated with a high morbidity and mortality. Although HAP is not a reportable illness, the available data indicate that it occurs at a rate of 5 to 10 cases per 1000 hospital admissions, and that this rate is 6- to 20-times higher in patients subjected to mechanical ventilation (3,4). Nevertheless, the incidence density of VAP varies widely depending on the case definition of pneumonia and the hospital population evaluated. Numbers of reported episodes per 1000 days of ventilation are: 34.5 after major heart surgery (5), 26 in a burns ICU (6), 18.7 in a pediatric ICU (7), 169

170

Bouza et al.

and between 8.0 (8) and 46.3 (9) in mixed medical/surgical ICUs. In the most recent report of the Centers for Disease Control National Nosocomial Infection Surveillance System it is stated that surgery and trauma ICUs have the highest VAP rates (mean 15.2/1000 ventilator days), followed by medical ICUs (mean VAP rate 4.9), coronary ICUs (mean VAP rate 4.4), and surgical ICUs (mean VAP rate 9.3) (10). The incidence of VAP in mechanically ventilated patients rises as the time of ventilation is prolonged. Early during the course of a hospital stay, the incidence of VAP is highest, with estimates of 3% per day during the first five days of ventilation, 2% per day from days 5 to 10, and 1% per day thereafter (11). Approximately half of all VAP episodes occur within the first four days of mechanical ventilation. The intubation process itself carries a risk of infection such that when acute respiratory failure is noninvasively managed, the rate of nosocomial pneumonia is lower (12–16). The overall mortality rate for HAP may be as high as 30% to 70%, but many critically ill patients with HAP die of their underlying disease rather than of pneumonia. VAP-related mortality has been estimated at 33% to 50% in several case-matched studies. Increased mortality rates have been attributed to the factors: bacteremia, especially that caused by Pseudomonas aeruginosa or Acinetobacter spp., medical rather than surgical illnesses, and ineffective antibiotic therapy (17–19). As well as being the leading cause of nosocomial mortality, VAP is the leading cause of nosocomial morbidity. Secondary bacteremia and empyema have been reported to occur in 4% to 38% and 5% to 8% of cases, respectively. On average, the hospital stay of VAP patients is extended for 4 to 13 days (median 7.6 days). Current estimates indicate that this additional length of stay generates a cost of $20,000 to $40,000 per case of HAP or VAP in the ICU.

PATHOGENESIS The pathogenesis of HAP and VAP is linked to two separate, but related, processes: colonization of the aerodigestive tract with pathogenic bacteria and aspiration of contaminated secretions. For VAP to occur, the delicate balance between host defenses and microbial invasion has to be upset, allowing pathogens to colonize the lower respiratory tract (20). In healthy subjects, the oropharynx is colonized by generally nonpathogenic microorganisms including Streptococcus viridans, Streptococcus pneumoniae, several anaerobes, and occasionally Haemophilus influenza, yet it is rare to find opportunistic gram-negative rods such as P. aeruginosa and Acinetobacter spp. Several factors have been reported to contribute to the pathogenesis of VAP such as the severity of the underlying disease, prior surgery, exposure to antibiotics, as well as the use of invasive respiratory equipment (2,21–30). Oropharyngeal and tracheal colonization by P. aeruginosa and enteric gram-negative bacilli have been related to the length of hospital stay and the severity of the underlying disease (26). The main route of VAP infection is oropharyngeal colonization by normal flora or by exogenous pathogens acquired in the ICU. Typical sources of these pathogens are the hands of medical staff or contaminated respiratory equipment, water, or air. Once the oropharynx has been invaded, microorganisms may reach the lower respiratory tract and lungs through several mechanisms. The main portals of bacterial entry into the lungs are oropharyngeal pathogen aspiration or the leakage of

Nosocomial Pneumonia in the Critical Care Unit

171

bacteria-containing secretions around the endotracheal cuff. The stomach and sinuses may act as potential reservoirs for nosocomial pathogens colonizing the oropharynx, but their relative role is largely unknown and could depend on the patient population or the changing natural history and management of VAP. Microaspiration is common even in healthy individuals. Approximately 45% of healthy subjects aspirate during sleep, and the rate of aspiration is higher in patients with reduced levels of consciousness. Factors promoting aspiration include a generally reduced level of consciousness, a diminished gag reflex, abnormal swallowing for any reason, delayed gastric emptying, or decreased gastrointestinal motility. Reflux and aspiration of nonsterile gastric contents is also a possible mechanism of pathogen entry into the lungs. The risk of pneumonia is determined by the type and quantity of bacteria colonizing the oropharynx (31). Hospitalized patients may become colonized with aerobic gram-negative bacteria within several days of admission, and as many as 75% of severely ill patients will be infected within 48 hours (32). In addition, the near sterility of the stomach and upper gastrointestinal tract may be disrupted by alterations in gastric pH due to illness, medication, or enteric feeding. Much attention has therefore been paid to the possible detrimental effects of ulcer prophylaxis regimens that raise the gastric pH (29,30). Orotracheal intubation diminishes the natural defense mechanisms of the respiratory tract affecting mechanical factors (ciliated epithelium and mucus), humoral factors (antibody and complement), and cellular factors (polymorphonuclear leukocytes, macrophages, lymphocytes, and their respective cytokines). The dorsal decubitus position is more conducive to microaspiration. The use of a nasogastric tube obstructs the ostia of the facial sinuses. The sinuses may then act as an infection reservoir causing the cardias to remain permanently open and inducing gastroesophagic reflux (33–35). The formation of a biofilm on the endotracheal tube could help sustain tracheal colonization, and this mechanism is also thought to play a role in late-onset VAP caused by resistant organisms. In summary, most cases of endemic VAP are acquired through the aspiration of microorganism-containing oropharyngeal, gastric, or tracheal secretions around the cuffed endotracheal tube into the normally sterile lower-respiratory tract. On the contrary, the most common source of epidemic VAP infection is contaminated respiratory treatment equipment, bronchoscopes, medical aerosols, water (e.g., Legionella), or air (e.g., Aspergillus). Direct inoculation of pathogens through ventilation devices is possible if no preventive measures are taken. Bacterial contamination of equipment was able to account for several VAP outbreaks in the 1970s, although today’s improved hygiene has meant that this route is only responsible for a few isolated outbreaks. Water condensing in the ventilation circuit is a potential source of contamination and several preventive measures are specifically recommended (see below) to avoid the risk of contamination via this route (2,22–25,27). The inhalation of pathogens such as viruses, fungi (Aspergillus spp.), or even Legionella spp. from the environment (2,15,22) has also been described. Pneumonia can also be acquired by the spread of infection from adjacent infected tissue such as the pleura or mediastinum, but this occurs very rarely. Bacterial translocation from the gastrointestinal tract is another pathogenic mechanism described for VAP. The intestinal wall of critically ill patients loses its capacity to prevent the systemic absorption of bacteria and toxins. This in turn leads

172

Bouza et al.

to impaired intestinal function, promoting the invasion of the blood system with intestinal pathogens and causing bacteremia and thus metastatic infections (36,37). The hematogenous spread of pathogens from intravascular catheters seems to be rare. An exception to the idea that ‘‘pathogenesis always starts with oropharyngeal colonization’’ is the case of infection by Pseudomonas spp. The findings of several studies indicate that tracheal colonization by these pathogens may occur without previous oropharyngeal colonization (38–40).

MICROBIOLOGY Although HAP and VAP are caused by a wide range of bacterial pathogens and some infections are polymicrobial [rates are especially high in patients with adult respiratory distress syndrome (ARDS)], viral or fungal pathogens are rarely the causative agents in immunocompetent patients. Sixty percent of nosocomial pneumonias are caused by gram-negative bacilli, representing six of the seven most frequently identified pathogens: P. aeruginosa (17%), Staphylococcus aureus (16%), Enterobacteriaceae (11%), Klebsiella spp. (7%), Escherichia coli (6%), H. influenza (6%) and Serratia marcescens (5%) (41). Moreover, in some hospitals, Acinetobacter spp. are starting to account for a significant number of cases of nosocomial pneumonia (42–44). Gram-positive cocci, such as S. aureus, particularly methicillin-resistant S. aureus (MRSA) represent another source of infection. The detection of an increased load of oropharyngeal commensals (viridans group streptococci, coagulase-negative staphylococci and Corynebacterium spp.) in distal bronchial specimens is difficult to interpret, but it is not generally considered that they could cause pneumonia. Rates of polymicrobial infection are highly variable although they seem to be on the increase and are particularly high in patients with ARDS (22,23,25,45–51). In a series of 104 patients over 75 years of age with severe pneumonia, El-Solh et al. identified S. aureus (29%), enteric gram-negative rods (15%), S. pneumoniae (9%), and Pseudomonas spp. (4%) as the pathogens most commonly responsible for nursing-home–acquired pneumonia (47). Generally, both nonventilated and ventilated patients show similar bacteriology and infection is usually provoked by multidrug-resistant pathogens (MDR) such as MRSA, P. aeruginosa, Acinetobacter spp., and K. pneumoniae. Pneumonia due to S. aureus is more common in patients with diabetes mellitus, head trauma, and ICU patients (4,47,52,53). P. aeruginosa is a frequent pathogen in patients with severe chronic obstructive pulmonary disease (COPD) and those with prior hospitalization, prolonged intubation (more than eight days), and prior exposure to antibiotics (54). Infection with Acinetobacter baumannii has been related to specific risk factors (55) including neurosurgery, ARDS, head trauma, and large-volume pulmonary aspiration. MDR-related VAP rates have recently undergone a dramatic increase in hospitalized patients. These pathogens are more likely to infect patients with late-onset HAP and VAP. The following risk factors for colonization and infection with MDR pathogens have been identified (2,19,56–59): 1. Antimicrobial therapy in the preceding 90 days 2. A length of hospital stay of five days or more 3. An existing high incidence of resistance to antibiotics in the hospital area or unit

Nosocomial Pneumonia in the Critical Care Unit

173

4. Risk factors for health-care–associated pneumonia a. Hospitalization for two days or more in the preceding 90 days b. Residence in a nursing home or extended care facility c. Home infusion therapy (including antibiotics) d. Chronic dialysis within the previous 30 days e. Home wound care f. Family member with a multidrug-resistant infection 5. Immunosuppressive disease and/or therapy Today, the role of anaerobic bacteria is still under investigation (60). In one report, anaerobes were isolated from 23% of patients with VAP diagnosed by quantitative culture methods (45). The authors of this study highlighted that the anaerobes recovered mirrored the bacteriology of the oropharynx and that only in four patients were they the only microorganism isolated. No anaerobic bacterium was found in blood or associated with necrotizing disease. In a more recent study, however, no anaerobes could be recovered using the same culture methods in 143 patients strictly followed during 185 episodes of VAP (61). Collectively, these and other findings point to an unlikely role of anaerobes in VAP or late-onset HAP. Their role in patients with poor dentition could, however, be more significant. Early-onset and late-onset disease can be distinguished using quantitative culture methods of diagnosis. When pneumonia develops within four or five days of admission (or intubation), microorganisms associated with community-acquired pneumonia are isolated with some frequency. In contrast, when disease develops after five days, few pathogens associated with community-acquired pneumonia are recovered and gram-negative bacilli and S. aureus are the main agents detected. Although indicators of late-onset disease, these bacteria can also cause early-onset pneumonia, especially in patients with severe comorbidities under recent antimicrobial treatment, making the distinction between early-onset and late-onset disease more difficult. As mentioned above, a longer period of mechanical ventilation and antimicrobial therapy will increase the risk of infection by MDR pathogens. Fungal pathogens are uncommon in immunocompetent hosts. Nosocomial Aspergillus spp. infection should warn of airborne transmission by spores related to an environmental source, such as contaminated hospital air ducts. Recently, a high rate of hospital-acquired Aspergillus pneumonia was observed in patients with COPD under therapy with antibiotics and high-dose corticosteroids (62). Candida albicans or other Candida species are often detected in endotracheal aspirates (EA), but usually indicate airway colonization rather than pneumonia and antifungal treatment is rarely necessary (63–67). Within the categories described, the causes of nosocomial pneumonia also vary considerably according to geographic, temporal, and intra-hospital factors. The use of up-to-date local epidemiologic ICU data on endemic pathogens can help select the most appropriate empirical antibiotic regimen and infection control strategies. Table 1 lists the conditions that may predispose a patient to acquire VAP attributable to a specific pathogen. HAP and VAP of viral cause is also rare in immunocompetent hosts. Outbreaks of HAP and VAP due to viruses, such as influenza, parainfluenza, adenovirus, measles, and respiratory syncitial virus have usually been seasonal. Influenza, parainfluenza, adenovirus, and respiratory syncitial virus account for 70% of all nosocomial viral pneumonias. The diagnosis of these viral infections is often made by rapid antigen testing and viral cultures or serological assays. Influenza A is probably

174

Bouza et al.

Table 1 Risk Factors for Ventilator-Associated Pneumonia Attributable to a Specific Microorganism Risk factor Aspiration Abdominal surgery IV drug abuse Coma Diabetes mellitus Chronic renal failure Prolonged ICU or hospital stay Antimicrobial therapy

Chronic obstructive pulmonary disease Age >65 yr Hypoalbuminemia 11,000 þ bands (>500) Abundant and purulent Localized 250 mmHg, and a normal white blood cell count is found in 73.3%, 74.7%, and 53.3% of patients, by this time, fever resolve is 73.3% of patients, the PaO2/ FiO2 is > ;250 mm Hg is 74.7% of patients and a normal white blood cell count is found in 53.3% of patients (54). Other authors report that infection variables resolve after antimicrobial therapy in patients with VAP by day 6 (203). Resolution of radiologic opacities and clearance of secretions occur at a median time of 14 and 6 days, respectively (54). However, failure to improve after 48 hours of therapy occurs in 65% of ARDS patients (54). Thus, ARDS significantly delays the clinical response to

184

Bouza et al.

Table 5 Performance of the Different Culture Methods for the Diagnosis of VentilatorAssociated Pneumonia Diagnostic technique Conventional Tracheal aspirates Tracheal aspirates BAL Protected specimen brush Plugged telescopic catheter Blind Tracheal aspirates Bronchial suction Mini BAL Protected specimen brush Protected telescopic catheter

Cutoff

Sensitivity

Specificity

References

105 cfu/mL

80% (60–97)

62% (41–74)

(196,207–209)

106 cfu/mL

66% (38–82)

78% (72–85)

(198,210,211)

104 cfu/mL 103 cfu/mL

73% (42–93) 66% (33–100)

82% (45–100) 90% (50–100)

(199,212–217) (215–217)

103 cfu/mL

72% (54–100)

82% (58–93)

(207,218,219)

105 cfu/mL

94%

50%

(220)

103–104 cfu/mL 103–104 cfu/mL 103 cfu/mL

74–97% 63–100% 66% (54–98)

74–100% 66–100% 91% (57–100)

(221) (220,221) (209,221,222)

103 cfu/mL

65%

83%

(220)

Note: Range given in parenthesis. Abbreviations: cfu, colony forming units; BAL, bronchoalveolar lavage.

treatment in critically ill patients with VAP, although temperature is still the earliest resolution variable in this group of patients. Reassessment is necessary in patients who show no clinical improvement by day 3, whereas for those showing a good response, it may be possible to design an abbreviated course of therapy (204). Prompt empirical therapy for all patients suspected of having VAP should be balanced with the need to limit antimicrobial misuse in ICUs. The reassessment of the patient’s situation on the basis of culture results is another major principle. In patients with positive cultures, therapy can be tailored in terms of quality and duration. In patients with negative cultures, the need to continue with antimicrobial drugs should be promptly reassessed. Discontinuation of antimicrobial agents is presently recommended in patients with a stable condition although in deteriorating or critically ill patients, it is difficult to make this decision. The Value of Surveillance Several research teams have addressed the issue of whether routine systematic surveillance of EA cultures may serve as a predictive diagnostic tool for VAP, although results have been contradictory (5,205,206). In a study performed at our center, the pathogens present in surveillance cultures taken prospectively on a twice-weekly basis did not correlate well with cultures obtained on diagnosis of VAP (5). Table 5 summarizes the performance of the different culture methods for the diagnosis of VAP.

Recommended AB

Linezolid or Vancomycin

Antipseudomonal cephalosporin or Antipseudomonal carbapenem or b-lactam/b-lactamase inhibitor plus Antipseudomonal fluorquinolone or Aminoglycoside

Combination AB therapy

15 mg/kg/12 hr IV

ciprofloxacin 400 mg/8 hr IV; levofloxacin 750 mg/day amikacin, 20 mg/kg/day, single dose gentamicin, 7 mg/kg/day, single dose tobramycin, 7 mg/kg/day, single dose 600 mg/12 hr IV

cefepime 1–2 g/8–12 hr IV; ceftazidime 2 g/8 hr IV impipenem 500 mg/6 hr or 1 g/8 hr IV; meropenem 1 g/8 hr 4.5 g/6 hr piperacillin-tazobactam

Dosing

2 g/day IV–IM 500 mg/day IV–PO 400 mg/day PO 750 mg/12 hr PO or 400 mg/12 hr IV 1.5–3 g/6 hr IV 1 g/day IV–IM

Dosing

Abbreviations: AB, antibiotic; MRSA, methicillin-resistant S. aureus; MDR, multidrug-resistant pathogens; MSSA, methicillin-sensitive S.aureus; ESBL, extended-spectrum beta-lactam mases.

MRSA

Pathogens above and P. aeruginosa K. pneumoniae ESBL Acinetobacter spp.

Potential pathogen

Risk factors for MDR pathogens, late-onset, and any disease severity

No known risk factors for MDR pathogens, early onset, and any disease severity S. pneumoniae Ceftriaxone or H. influenzae Levofloxacin MSSA Moxifloxacin or Antibiotic sensitive gram-negative bacilli (E. coli, K. pneumoniae, Ciprofloxacin or Enterobacter spp., Proteus spp., S. marcescens) Ampicillin/sulbactam or Ertapenem

Potential pathogen

Table 6 Initial Empirical Antibiotic Treatment for Ventilator-Associated Pneumonia According to the Potential Pathogens

Nosocomial Pneumonia in the Critical Care Unit 185

186

Bouza et al.

ANTIMICROBIAL TREATMENT Selecting an Empirical Regimen When trying to overcome severe infection, cardiovascular support and measures to improve hemodynamics and oxygenation are critical (54). The most important lesson learned in the last decade on the management of VAP is probably that delaying effective antimicrobial therapy in these patients increases mortality (50,108,111,223), length of stay, and costs (224). As soon as there is clinical suspicion of VAP, adequate antibiotics should be administered to increase the likelihood of an early reduction in the bacterial load. The first step is to decide whether a patient carries a low or high risk of having a MDR pathogen. The main risk factors for a MDR pathogen are (i) five or more days of prior hospitalization or mechanical ventilation, (ii) exposure to antibiotics in the preceding 90 days, (iii) a high incidence of antimicrobial resistance in the specific hospital unit, and (iv) comorbidities such as use of corticosteroids, head trauma, and lung structural disease among others (20,55–57,59,225–229). Patients with none of these risk factors can start therapy with reducedspectrum drugs such as ceftriaxone, a fluorquinolone (levofloxacin, moxifloxacin), ampicillin/sulbactam, or ertapenem. If the patient has any of the risk factors for a MDR pathogen then a two to three drug regimen should be started, including an anti-Pseudomonas beta-lactam agent (cefepime or ceftazidime, or piperacillin/ tazobactam or imipenem or meropenem), a second anti-Pseudomonas agent (aminoglycoside or anti-Pseudomonas fluoroquinolone such as ciprofloxacin or levofloxacin) plus a broad-spectrum agent against gram-positive microorganisms (linezolid or vancomycin) (Table 6). Treatment should be started immediately after obtaining adequate samples for microbiological diagnosis. Treatment Based on Knowledge of the Etiologic Microorganism A key issue in the antimicrobial treatment of VAP is the de-escalation of treatment once microbiological information becomes available. We have already mentioned that antimicrobial agents should be discontinued when appropriate culture results are negative. Once 24 to 48 hours has passed, information on the number and type of microorganisms growing in culture should be available. Depending on whether there is a lack of gram-negative microorganisms or one of gram-positive microorganisms, the specific drug against the corresponding microorganisms can be withdrawn even before the identity and susceptibility of the etiologic agent is known. The microorganisms that deserve most attention are MRSA, P. aeruginosa, and A. baumannii. Vancomycin is presently the standard agent against MRSA, although both industry-sponsored clinical trials and data from individual centers have consistently reported clinical failure rates of 40% or greater, at least using a standard dose. New evidence suggests that vancomycin failure could be related to inadequate dosing (230,231) and some authors argue that trough levels of around 15 mg/L are needed (232), although the success of this strategy requires confirmation in clinical trials. The addition of rifampin, aminoglycosides, or other drugs has achieved little improvement (233). The use of new antimicrobial agents against MRSA has also been explored. Thus, quinupristin-dalfopristin has generated worse results than vancomycin (230).

Nosocomial Pneumonia in the Critical Care Unit

187

Linezolid, an oxazolidinone antimicrobial agent, is active against MRSA and achieves better tissue penetration than vancomycin, but is bacteriostatic rather than bactericidal (234,235). However, a combined analysis of the results of two randomized trials comparing linezolid with vancomycin for the treatment of nosocomial pneumonia (each in combination with aztreonam for gram-negative coverage) suggests a therapeutic advantage for linezolid (236). In a further analysis of a subset of patients with MRSA VAP, linezolid was associated with a significantly higher probability of bacterial eradication, clinical cure, and hospital survival (237). Despite higher costs, linezolid therapy for MRSA VAP was attributed an absolute mortality benefit of 22%, which translates into five patients as the number-needed-to-treat to save one life (237). This has led linezolid to become recommended therapy for MRSA VAP (20). Further agents presently under investigation include tigecycline, a new glycylglycine antimicrobial derived from tetracyclines. Tigecycline has an extremely broad spectrum of action against gram-positive, gram-negative, and anaerobic pathogens, with the exception of Pseudomonas (238). Its role in VAP is currently being evaluated in a phase III clinical trial. Pneumonia due to P. aeruginosa in ventilated patients is frequently a recurrent disease caused most of the time by several relapsing infections (239). Frequently, the pathogens are MDR such that no single antibiotic is active against all isolates. Empirical therapy includes the combination of two drugs active against P. aeruginosa to improve the chances of successful early treatment. Once the susceptibility pattern is known, many physicians prefer combination therapy with a beta-lactam agent plus either an aminoglycoside or an anti-Pseudomonas fluoroquinolone, based on early findings related to patients with bloodstream infections (240). There is presently, however, no evidence that combination therapy has any benefit over monotherapy in patients with VAP or other forms of nosocomial pneumonia (241–243). Moreover, the combined regime was even found to fail at avoiding the development of resistance during therapy (242). In select patients with infections caused by MDR strains, aerosolized colistin has proved beneficial as supplemental therapy (244). A. baumannii is a nonfermenter gram-negative rod, which has been held responsible for the recent rise in VAP. It is intrinsically resistant to many antimicrobial agents, and the agents found to be most active against them are the carbapenems, sulbactam, and polymyxins (42,54). In effect, intravenous carbapenem is today the treatment of choice for MDR isolates (245). In patients with strains resistant to carbapenems, intravenous colistin has been successfully used (43). Adequate Dosing To ensure the best outcome for patients, it is essential that the dosing of initial antibiotics for suspected MDR pathogens is adequate (235). All too often, agents are initially underdosed. For example, vancomycin should not be routinely given at a dose of 1 g q12h, but rather the dose should be calculated by weight in mg/kg (a dose that needs adjusting for renal impairment). Retrospective pharmacokinetic modeling has suggested that the failures described for vancomycin could be the result of inadequate dosing. Many physicians aim for a trough vancomycin concentration of at least 15 mg/L, although, as mentioned in the previous section, the success of this strategy has not been prospectively confirmed. Some antibiotics penetrate well and achieve high local concentrations in the lungs, whereas others do not. For example, most beta-lactam antibiotics achieve

188

Bouza et al.

less than 50% of their serum concentration in the lungs, whereas fluoroquinolones and linezolid attain equivalent or higher concentrations than blood levels in bronchial secretions. Table 7 shows how to adjust the antibiotic dose in patients with renal impairment.

Aerosolized Antibiotics All patients with VAP should initially receive antibiotics intravenously, but conversion to oral/enteral therapy may be possible in certain responding patients. The direct aerosol delivery of antibiotics is not considered as a standard therapy either for prophylaxis or for the treatment of lower respiratory tract infections (246). In the past, aminoglycosides and polymyxins were the most common agents used in aerosols. In a prospective randomized trial, the use of intravenous therapy was compared to the same treatment plus aerosolized tobramycin. The results of this trial suggest no better clinical outcome, but bacterial cultures of the lower respiratory tract were more rapidly eradicated (246b). At present aerosolized antimicrobial therapy is mainly limited to MDR pathogens for which no other treatment exists, such is the case of MDR P. aeruginosa and A. baumannii, which are treated with intratracheal colistin (244). Monotherapy or Combination Therapy When considering the use of a single antimicrobial agent or combined therapy, we first need to make the distinction between the use of multiple antimicrobial agents in the initial empirical regimen (to ensure that a highly resistant pathogen is covered by at least one drug) and that of combination therapy continued intentionally after the pathogen is known to be susceptible to both agents. The former use of combination therapy is uniformly recommended, whereas the latter use remains controversial. Two meta-analyses have recently explored the value of combination antimicrobial therapy in patients with sepsis (243) and gram-negative bacteremia (247). No benefits of combination therapy were shown, and nephrotoxicity in patients with sepsis or bacteremia increased. However, in the subset of bacteremic patients infected with P. aeruginosa, combination therapy (usually a beta-lactam and an aminoglycoside) reduced the risk of mortality by half. A trend towards improved survival has been previously observed with aminoglycoside-including, but not with quinolone-including combinations (8). Combination therapy could therefore be beneficial in patients with severe, antimicrobial-resistant infections. Whether this benefit is due to more reliable initial coverage or to a synergistic effect is unclear. The present general consensus is to use combination therapy with an aminoglycoside for the initial five days in patients with VAP caused by gram-negative bacilli (20,162). However, the nephrotoxicity of aminoglycosides limits the use of these agents. Duration of Therapy The ideal length of antibiotic therapy is still under debate. In a prospective randomized clinical trial, Chastre et al. (248) demonstrated that an eight-day antibiotic regimen is comparable to a 15-day regimen in terms of mortality, superinfections, and relapse of VAP. A seven-day treatment course was described as safe, effective,

Nosocomial Pneumonia in the Critical Care Unit

189

Table 7 Antibiotic Dose Adjustment in Renal Impairment Antibiotic Amicacin

Ampicillin/ sulbactam Cefepime

Ceftazidime

Ceftriaxone Ciprofloxacin

Ertapenem Gentamicin

Imipenem

CrCl (mL/min)

Dose adjustment

40 30–39 20–29 30 15–30 60 30–60 11–29 50 10–50 50 750 mg/12 hrs PO 400 mg/12 hrs IV 10–50 250–500 mg/12 hrs PO 400 mg/18 hrs IV 31 No adjustment 30 500 mg/24 hrs IV–IM 50 5 mg/kg/24 hrs 30–49 5 mg/kg/36 hrs 20–29 5 mg/kg/48 hrs 50 20–49 50 26–50 10–25 40 20–40 2.5 g/dL WBC > 500/mm3 Fluid is serous or cloudy, free flowing Fibropurulent stage Thick fluid and positive cultures Organizing stage Organizing peel with entrapped lung

Drainage Thoracentesis þ antibiotics

Large-bore thoracostomy tube  fibrinolytics Open thoracostomy Mini-thoracotomy Thoracoscopic evacuation Decortication

Abbreviations: LDH, lactate dehydrogenase; WBC, white blood count.

210

Bartlett

The fibropurulent phase is associated with fluid that is too thick for drainage by thoracentesis necessitating thoracostomy, often with large-bore needles. The drainage may be facilitated by catheter placement under fluoroscopic, computed tomographic, or ultrasonic guidance (22,23,51,52). Closed drainage with suction is recommended if the fluid is thick, there is evidence of a bronchopleural fistula, or the pleural fluid is putrid. The thoracostomy tubes are left in place until the cavity is obliterated, and the yield of pleural drainage is less than 25 mL/day, any bronchopleural fistula is sealed, and fever has resolved. Failure to achieve these goals may indicate the requirement for chest tube positioning or reinsertion. Failure of the closed procedure often indicates the need for open drainage with rib resection or decortication. Decisions regarding drainage procedure may be facilitated by the empyema severity score based on pleural fluid pH glucose and results of imaging—ultrasonography to localize loculated pleural fluid prior to aspiration or thoracostomy, or CT scan to detect a pleural peel (56). Factors indicating a possible need for surgery are failure to defeverse with antibiotics and thoracentesis and a severity of pleural disease score based on low pleural fluid pH, low glucose, moderate or severe scoliosis, presence of pleural peel, and infection due to anaerobes or gram-negative bacilli. The initial surgical procedure is often a closed tube thoracostomy using ultrasound or CT guidance for tube placement. Fibrolytic agents are often instilled through the tubes, but their utility is not clearly established. The next stage for patients who do not respond is rib resection with open drainage, a procedure associated with substantial morbidity. A variant is the mini-thoracotomy, which involves a 5 cm incision and short segment rib resection (57). The late organizing phase is characterized by an extensive collection of fibrous material or pleural peel and lung entrapment, which generally requires open thoracotomy or decortication (9,41,48). This procedure is done with open thoracotomy, debridement of the coagulum, careful excision of the peel, and assessment of lung expansion prior to closure. A variant recently introduced is decortication with video assistance to reduce morbidity. Outcome Studies from the prepenicillin era showed mortality rates of 7% to 41% with pneumococcal empyema (10,11). The mortality is obviously much lower in the current era associated with antibiotics and improved thoracic surgery but still reported at 8% to 20% (3–12,58,59). This includes studies published since 1984 (3,60). Factors that indicate a poor prognosis include chronic empyema, nosocomial empyema, advanced age, association with malignancy, and the presence of a bronchopleural fistula. Excluding patients with serious or ultimately lethal underlying conditions, the mortality rates are generally 3% to 6% (3,57). In many of these cases, the major difficulty is achieving adequate drainage. These cases are associated with long hospital stays and continuing controversies about the next step in the patient with persistent pleural collections. LUNG ABSCESS Introduction Lung abscess results from necrosis of lung parenchyma. The term ‘‘necrotizing pneumonia’’ and ‘‘pulmonary gangrene’’ are sometimes used to indicate small pulmonary abscesses in contiguous areas of the lung. These represent a continuum and are

Pleural Empyema/Lung Abscess in the Critical Care Unit

211

infections involving a relatively small number of pathogens that can cause this pathologic result. Thus, as with empyema, the tabulation of microbes that cause pneumonia is legion; the number that cause pulmonary necrosis is a relatively short list. Classification A number of methods to classify lung abscess and either the terms or the definitions have been used in the literature. Some of the standard terms that have been used during the past 50 years include  Acute or chronic abscesses with four to six weeks as the standard dividing line.  Primary versus secondary abscesses; the former generally refers to abscesses in patients prone to aspiration or in healthy adults while secondary abscesses represent complications of a primary condition of the lung such as a neoplasm or HIV infection.  Nonspecific lung abscess often refers to abscesses in which no likely pathogen is recovered from expectorated sputum; most of these are considered anaerobic infections.  Putrid lung abscess indicates the offensive odor of sputum that is considered diagnostic of anaerobic infection.  Nosocomial lung abscess in reference to those that occur during hospitalization and are usually due to nosocomial pathogens. In an extensive experience with over 1000 reported cases of lung abscess in the antibiotic era, approximately 80% were considered primary, 60% putrid, 40% nonspecific, and 40% chronic (61–68). Clinical Features Lung abscesses may be acute or chronic. Chronic abscesses are most likely to represent microbacteria infections due to mycobacteria, anaerobic bacteria, and melioidosis. The latter is a relatively common cause of pulmonary infections due to Burkholderia pseudomallei in Asia; this is extremely rare in the United States except for occasional immigrants. Acute lung abscess is usually associated with more virulent organisms such as Klebsiella and S. aureus. The usual clinical features of lung abscess include fever, fatigue, cough with sputum production, and sometimes pleurisy and/or hemoptysis. Chronic lung abscesses are usually accompanied by weight loss and anemia. Approximately 60% of patients with bacteriologically confirmed lung abscesses that are due to anaerobic bacteria have putrid sputum, empyema fluid, or breath. Evaluation The characteristic feature on chest X-ray is a cavity in the pulmonary parenchyma often with an air-fluid level within a pulmonary infiltrate. Lymphadenopathy suggests some specific conditions such as mycobacterial or fungal infection. CT is a particularly sensitive method to detect lung abscess and provides good anatomic definition. This also will clearly distinguish air-fluid levels in the pleural space from those within the pulmonary parenchyma (68,69). Segmental location of the abscess is important in the differential diagnosis. Anaerobic lung abscess due to aspiration

212

Bartlett

Table 3 Pulmonary Lesions with Appearance of Lung Abscess Necrotizing infections of the lung Bacteria: anaerobic bacteria, microaerophilic/anaerobic streptococci (Streptococcus milleri), enteric GNB (esp. Klebsiella), Pseudomonas aeruginosa, Staphylococcus aureus (including MRSA USA 300) Less common: nocardia, actinomycosis, Rhodococcus, Legionella, Pasteurella multocida Mycobacteria: Mycobacterium tuberculosis, M. kansasii, M. avium-intracellulare Fungi: Coccidioidomycosis, Histoplasmosis, Blastomycosis, Aspergillus, Cryptococcus, Mucor Parasite: Entamoeba histolytica, Paragonimus westermani, Echinococcus Cavity Bland infarct  infection Septic emboli Vasculitis: Wegner’s granulomatosis, periarteritis nodosa Neoplasms: lung cancer, metastatic carcinoma, lymphoma Others: cysts, blebs, pneumatocele, sequestration, bronchiectasis, empyema Abbreviation: MRSA, methicillin-resistant Staphylococcus aureus.

usually involves dependent pulmonary segments, the posterior segments of the upper lobe of superior segments of the lower lobes. These reflect the dependent pulmonary segments in the recumbent position. Tuberculosis favors the upper lobes. There are a number of conditions to consider in the patient with an established or suspected abscess involving the pulmonary parenchyma as indicated in Table 3. Detection of the etiologic agent is particularly important because these infections are often both severe and chronic so that pathogen-specific therapy is highly desired. A good pretreatment expectorated sputum is useful for Gram stain and culture; the diagnostic yield is generally excellent for detection of S. aureus and aerobic gram-negative bacilli, particularly Klebsiella pneumoniae or Pseudomonas aeruginosa. This specimen is not appropriate for anaerobic culture so that these organisms, the most common causes of lung abscess, must be treated empirically or there needs to be a diagnostic method which bypasses the contamination associated with specimens that traverse the upper airways. In former years this was done by transtracheal aspiration (TTA), which became the technique to define the etiology of aspiration pneumonia and lung abscess in the 1970s (58,59,70,71). The TTA has subsequently been abandoned. More recent studies have sometimes used quantitative cultures of bronchoscopic specimens, either bronchoalveolar lavage or use of the protected brush (72–75). The experience with this technique is quite variable, in part related to the need for specimens prior to antibiotic treatment, the deleterious effect of lidocaine on some fastidious anaerobes, the difficulty that many laboratories have in cultivating oxygen-sensitive bacteria, and inconsistent methods of specimen collection and processing. Thus, many authorities do not regard this as an appropriate method to either rule in or rule out anaerobic bacteria. An alternative method that is occasionally used is transthoracic aspiration in which there is needle aspiration of the abscess percutaneously. Bacteriology Anaerobic Bacteria These are the organisms that represent normal flora of the oral cavity, primarily the gingiva crevice. They reach the lung by aspiration, usually in a host that is prone to

Pleural Empyema/Lung Abscess in the Critical Care Unit

213

aspiration due to decreased consciousness or dysphagia. The dominant isolates in these cases are Peptostreptococcus spp., Prevotella melaninogenica, and Fusobacterium nucleatum (58,59,70,71,76,77). Staphylococcus aureus This organism has been historically found primarily in pediatric patients, patients with influenza complicated by bacterial superinfections, injection drug users with tricuspid valve endocarditis with septic emboli to the lung, and with nosocomial pulmonary infections (78,79). More recently there has been a new syndrome attributed to ‘‘CA-MRSA,’’ an epidemic strain referred to as ‘‘USA300’’ (80–84). The latter actually refers to a family of related strains, but those involved are a relatively homogeneous group that is characterized by the presence of genes for the Paton-Valentine Leukocidin, a possible virulence factor, and the mecIV element, which confers resistance to all betalactams (85). Many or most of these strains are sensitive to many antibiotics that are not generally found to be active in vitro with nosocomial strains of MRSA; these include trimethoprim-sulfamethoxazole, gentamicin, rifampin, and doxycycline. The clinical syndrome associated with this pathogen is usually devastating. Most of the patients are young, previously healthy adults, who acquire influenza and then have a rapid and fulminant course characterized by necrotizing pneumonia and shock. The putative is recovered from blood and/or respiratory secretions. The mortality rate is variously reported at 20% to 50% (80–84). Klebsiella Klebsiella has always been recognized as a possible pulmonary pathogen with a penchant to cause abscess (86,87). The classic description was in the prepenicillin era when the typical presentation was a debilitated host who presented with fever, cough, pleurisy, and sputum that looked like currant jelly and an X ray that showed an upper lobe infiltrate with the ‘‘bulging fissure sign’’; this went on to cavitate. This form of Klebsiella pulmonary infection is rarely seen currently, but Klebsiella continues to be occasionally implicated in lung abscess, particularly in Taiwan where it is epidemic (87). Miscellaneous Bacteria Other agents implicated include nocardia (88,89) Legionella (90–92), Streptococcus milleri (93,94), S. pyogenes (95,96), Rhodococcus equi (97–100), Fusobacterium necrophorum (as a complication of Lemierre syndrome) (101–103), and B. pseudomallei (melioidosis) (104).

Treatment Most patients respond to antibiotic treatment, and this is best if it is pathogenspecific. Tuberculosis is treated with the standard four drug regimen. Klebsiella generally responds to cephalosporin treatment, but in vitro sensitivity tests are desired, and P. aeruginosa is found primarily in the compromised host or the patient with structural disease of the lung; optimal therapy is controversial due to the debate concerning the need for one or two drugs directed against this pathogen (105,106). Long courses of a fluoroquinolone selected on the basis of in vitro activity in patients

214

Bartlett

with HIV infection show high rates of failure, suggesting that alternative drugs may be preferred (106). For infections involving S. aureus (USA300), most are MRSA, and the recommended treatment is either vancomycin combined with rifampin or linezolid combined with rifampin (25,26). Some strains are methicillin-sensitive and should be treated with oxycillin or nafcillin. Patients tend to do poorly despite seemingly adequate antibiotic treatment (80–84). It is possible that antibiotics directed against exotoxin production such as clindamycin or linezolid are important (81). For abscesses involving anaerobic bacteria, penicillin and tetracyclines were the standard drugs in the 1950s and 1960s when these were referred to as nonspecific lung abscesses; most patients responded. More recent trials have shown that clindamycin is superior to penicillin in terms of time to defervescence, time to eradication of putrid sputum and overall response rates (107,108). Any betalactam–betalactamase inhibitor combination also appears to be successful in the great majority of cases (109,110). Metronidazole is active against virtually all anaerobic bacteria but has a relatively poor track record in the treatment of anaerobic lung infections (111–113). The presumed explanation is the potentially important role of microaerophilic and aerobic streptococci in these infections. This may be obviated by the addition of penicillin to metronidazole. The duration of treatment is arbitrary, and many patients relapse when the treatment is discontinued prematurely. Current recommendations are to treat for six weeks or, preferably, treat until the X ray is clear or shows only a small, stable residual scar. Drainage is generally an essential component of managing abscesses, but this does not usually apply to lung abscess, which drains spontaneously by communication with the bronchus. There is an occasional attempt to facilitate drainage with physical therapy for postural drainage or bronchoscopy. Studies in the prepenicillin era showed bronchoscopic therapy had virtually no effect on outcome. For patients who fail to respond, thoracic surgery is occasionally required, usually in about 5% to 10% (114–117). The usual indications are failure of medical management and suspected neoplasm or hemorrhage. Failure to respond is usually associated with an extremely large abscess measuring greater than 6 cm in diameter, abscesses that have been present for prolonged periods, and those due to relatively resistant organisms such as P. aeruginosa. When surgery is performed, the usual procedure is a lobectomy or pneumonectomy. An alternative drainage option is drainage under computed tomographic guidance which has a modest but favorable reported experience. Outcome Patients with lung abscess due to anaerobic bacteria usually show decrease in fever within three to four days of initiating antibiotic treatment with complete defervescence over 7 to 10 days (27,62,64,107). Patients with fever lasting more than 7 to 14 days usually undergo bronchoscopy or other diagnostic intervention to better define the microbiology. CT scans may be particularly useful for anatomic definition to determine possible obstruction or adenopathy that would prompt an investigation for alternative pathogens such as TB or fungi. Mortality rates for lung abscess are usually reported at 5% to 15% but are much lower if there is no major underlying, ultimately fatal condition. A review of lung abscess cases in Japan showed a mortality of 2% in community-acquired cases and 67% for nosocomial infections (65).

Pleural Empyema/Lung Abscess in the Critical Care Unit

215

REFERENCES 1. Bartlett JG. Bacterial infections of the pleural space. Semin Respir Infect 1988; 3:309. 2. Light RW. A new classification of Parapneumonic effusions and empyema. Chest 1995; 108:299. 3. Maskell NA, Davies CW, Nunn AJ, et al. U.K. Controlled trial of intrapleural streptokinase for pleural infection. N Engl J Med 2005; 352:865–874. 4. Mayo P. Early thoracotomy and decortication for nontuberculous empyema in adults with and without underlying disease: a 25-year review. Am Surg 1985; 51:230. 5. Benfield G. Recent trends in empyema thoracis. Br J Dis Chest 1981; 75:358. 6. Alfageme I, Munoz F, Pena N, Umbria S. Empyema of the thorax in adults: etiology, microbiologic findings, and management. Chest 1993; 103:839. 7. Wiedemann HP, Rice TW. Lung abscess and empyema. Semin Thorac Cardiovasc Surg 1995; 7:119. 8. LeMense GP, Strenge C, Sahn SA. Empyema thoracis: therapeutic management and outcome. Chest 1995; 107:1532. 9. deSousa A, Offner PJ, Moore EE, et al. Optimal management of complicated empyema. Am J Surg 2000; 180:507. 10. Finland M. The significance of pneumococcal types in disease, including types IV to XXXII (Cooper). Ann Intern Med 1939; 15:1531. 11. Heffron R. Pneumonia. Cambridge, MA: Harvard University Press, 1939:566–585. 12. Keefer C, Rantz L, Rammelkamp C. Hemolytic streptococcal pneumonia and empyema: a study of 55 cases with special reference to treatment. Ann Intern Med 1941; 14:1533. 13. Welch C, Tombridge T, Baker W, et al. Beta-hemolytic streptococcal pneumonia: report of an outbreak in a military population. Am J Med Sci 1961; 242:157. 14. Fine MJ, Smith MA, Carson CA, et al. Prognosis and outcomes of patients with communityacquired pneumonia. JAMA 1996; 275:134. 15. Deschamps C, Allen MS, Trastek VA, Pairolero PC. Empyema following pulmonary resection. Chest Surg Clin North Am 1994; 4:583. 16. Bernard A, Pillet M, Goudet P, Viard H. Antibiotic prophylaxis in pulmonary surgery: a prospective randomized double-blind trial of flash Cefuroxime versus forty-eight-hour Cefuroxime. J Thorac Cardiovasc Surg 1994; 107:896. 17. Heffner JE, Brown LK, Barbieri C, DeLeo JM. Pleural fluid chemical analysis in Parapneumonic effusions: a meta-analysis. Am J Respir Crit Care Med 1995; 151:1700. 18. Hiraki A, Aoe K, Eda R, et al. Comparison of six biological markers for the diagnosis of tuberculous pleuritis. Chest 2005; 125:987–989. 19. Diacon AH, Van de Wal BW, Wyser C, Smedema JP, et al. Diagnostic tools in tuberculous pleurisy: a direct comparative study. Eur Respir J 2003; 22:589–591. 20. Chapman SJ, Davies RJ. The management of pleural space infections. Respirology 2004; 9:4–11. 21. Villena V, Lopez-Encuentra A, Garcia-Lujan R, Echave-Sustaeta J, Martinez CJ. Clinical implications of appearance of pleural fluid at thoracentesis. Chest 2004; 125:156–159. 22. Yim AP, Ho JK, Lee TW, Chung SS. Thoracoscopic management of pleural effusions revisited. Aust NZ J Surg 1995; 65:308. 23. Hunnam GR, Flower CD. Radiologically-guided percutaneous catheter drainage of empyemas. Clin Radiol 1988; 39:121. 24. Sullivan K, O’Toole R, Fisher R, et al. Anaerobic empyema thoracis. Arch Intern Med 1973; 131:521. 25. Bartlett JG, Thadepalli H, Gorbach S, et al. Bacteriology of empyema. Lancet 1974; 1:338. 26. Porcel JM, Vives M. Etiology and pleural fluid characteristics of large and massive effusions. Chest 2003; 124:978–983. 27. Bartlett JG. Anaerobic bacterial infections of the lung and pleural space. Clin Infect Dis 1993; 16:S248.

216

Bartlett

28. Beerens H, Tahon-Castel M. Infection humaines a bacteries anarobes non-toxigenes. Brussels: Academiques Europeennes, 1965:92. 29. Kelly JW, Morris MJ. Empyema thoracis: medical aspects of evaluation and treatment. South Med J 1994; 87:1102. 30. Inglesby TV, O’Toole T, Henderson DA, et al. Anthrax as a biological weapon, 2002: updated recommendations for management. JAMA 2002; 288:1849. 31. Jernigan JA, Stephens DS, Ashford DA, et al. Anthrax Bioterrorism Investigation Team. Bioterrorism-related inhalational anthrax: the first 10 cases reported in the United States. Emerg Infect Dis 2001; 7:933–944. 32. Kyriacou DN, Stein AC, Yarnold PR, et al. Clinical predictors of bioterrorism-related inhalational anthrax. Lancet 2004; 364:449–452. 33. Sokolowski JW Jr., Burgher LW, Jones FL Jr., et al. Guidelines for thoracentesis and needle biopsy of the pleura. Am Rev Respir Infect 1988; 140:257. 34. Schabel SL. Imaging of pleural infections. Semin Respir Infect 1988; 3:298. 35. Levin DL, Klein JS. Imaging techniques for pleural space infections. Semin Respir Infect 1999; 14:31. 36. Teixeira LR, Sasse SA, Villarino MA, et al. Antibiotic levels in empyema fluid. Chest 2000; 117:1734. 37. Hughes CE, Van Scoy RE. Antibiotic therapy of pleural empyema. Semin Respir Infect 1991; 6:94. 38. Diacon AH, Theron J, Schuurmans MN, et al. Intrapleural streptokinase for empyema and complicated parapneumonic effusions. Am J Respir Crit Care Med 2004; 170:49. 39. Aye RW, Froese DP, Hill LD. Use of purified streptokinase in empyema and hemothorax. Am J Surg 1991; 37:371–374. 40. Light RW, Nguyen T, Mulligan ME, Sasse SA. The in vitro efficacy of varidase versus streptokinase or urokinase for liquefying purulent exudative material from locutated empyema. Lung 2000; 178:13. 41. Henke CA, Leatherman JW. Intrapleurally administered streptokinase in the treatment of acute loculated nonpurulent Parapneumonic effusions. Am Rev Respir Dis 1992; 45:680. 42. Pollak JS, Passik CS. Intrapleural urokinase in the treatment of loculated pleural effusions. Chest 1994; 105:868. 43. Miller JI Jr. The history of surgery of empyema, thora coplasty, eloesser flap and muscle flap transposition. Chest Surg Clin North Am 2000; 10:45. 44. Kearney SE, Davies CW, Davies RJ, Gleeson kFV. Computed tomography and ultrasound in Parapneumonic effusions and empyema. Clin Radiol 2000; 55:542. 45. Sasse S, Nguyen T, Teixeira LR, Light R. The utility of daily therapeutic thoracentesis for treatment of early empyema. Chest 1999; 116:1703–1708. 46. Light RW, Rodriguez RM. Management of Parapneumonic effusions. Clin Chest Med 1998; 19:373–382. 47. Powell LL, Allen R, Brenner M, et al. Improved patient outcome after surgical treatment for loculated empyema. Am J Surg 2000; 179:1. 48. Chen LE, Langer JC, Dillon PA, et al. Management of late-stage Parapneumonic empyema. J Pediatr Surg 2002; 37:371–374. 49. Colice GL, Curtis A, Deslauriers J, et al. Medical and surgical treatment of Parapneumonic effusions: and evidence-based guideline. Chest 2001; 119:319. 50. Manuel Porcel J, Vives M, Esquerda A, Ruiz A. Usefulness of the British Thoracic Society and the American College of Chest Physicians guidelines in predicting pleural drainage of non-purulent parapneumonic effusions. Respir Med 2005 [Epub ahead of print]. 51. Balfour-Lyn IM, Abrahamson A, Cohen G, et al. BTS guidelines for the management of pleural infection in children. Thorax 2005; 60:i1–i21. 52. O’Moore PV, Mueller PR, Simeone JF, et al. Sonographic guidance in diagnostic and therapeutic interventions in the pleural space. AJR 1987; 149:1.

Pleural Empyema/Lung Abscess in the Critical Care Unit

217

53. Davies CH, Gleeson FV, Davies RJO. BTS guidelines on the management of pleural infection. Thorax 2003; 58:ii18–ii28. 54. Maskell NA, Davies CWH, Ghabe R, et al. Predictors of survival in patients with pleural infection but without cancer: results from MRC/BTS MIST Trial, ICTN 39138989. Thorax 2004; 59:ii40. 55. Cantin L, Chartrand-Lefebvre C, Lepanto L, et al. Chest tube drainage under radiological guidance for pleural effusion and Pneumothorax in a tertiary care university teaching hospital: review of 51 cases. Can Respir J 2005; 12:209–233. 56. Hoff SJ, Neblett WW III, Heller RM, et al. Parapneumonic empyema in childhood: selecting appropriate therapy. J Pediatr Surg 1989:659–663. 57. Raffensperger JG, Luck SR, Shkolnik A, Ricketts RR. Mini-thoracotomy and chest tube insertion for children with empyema. J Thorac Cardiovasc Surg 1982; 84:497–504. 58. Bartlett JG, Rosenblatt JE, Finegold SM. Percutaneous transtracheal aspiration in the diagnosis of anaerobic pulmonary infection. Ann Intern Med 1973; 79:535. 59. Bartlett JG, Gorbach SL, Tally FP, et al. Bacteriology and treatment of primary lung abscess. Am Rev Respir Dis 1974; 109:510. 60. Jess P, Brynitz S, Friis Moller A. Mortality in thoracic empyema. Scand J Thorac Cardiovasc Surg 1984; 18:85–87. 61. Hagan J, Hardy JD. Lung abscess revisited: a survey of 184 cases. Ann Surg 1983; 197:755. 62. Harber P, Terry PB. Fatal lung abscesses: review of 11 years’ experience. South Med J 1981; 74:281. 63. Pohlson EC, McNamara J, Char C, Kurata B. Lung abscess: a changing pattern of the disease. Ann Surg 1985; 150:97. 64. Hirshberg B, Skair-Levi M, Nir-Paz R, et al. Factors predicting mortality of patients with lung abscess. Chest 1999; 115:746. 65. Mori T, Eb T, Takahashi M, et al. Lung abscess: analysis of 66 cases from 1979 to 1991. Intern Med 1993; 32:278. 66. Hammond JM, Potgieter PD, Hanslo D, et al. The etiology and antimicrobial susceptibility patterns of microorganisms in acute community-acquired lung abscess. Chest 1995; 108:937. 67. Bartlett JG. The role of anaerobic bacteria in lung abscess. Clin Infect Dis 2005; 40:923–925. 68. Landay MJ, Christensen EE, Bynum LJ, et al. Anaerobic pleural and pulmonary infections. AJR 1980; 134:233. 69. Stark DD, Federle MP, Goodman PC, et al. Differentiating lung abscess and empyema: radiography and computed tomography. AJR 1983; 141:163. 70. Bartlett JG. Diagnostic accuracy of transtracheal aspiration bacteriology. Am Rev Respir Dis 1977; 115:777. 71. Bartlett JG. The technique of transtracheal aspiration. J Crit Illness 1986; 1:43. 72. Wimberley NW, Bass JB, Boyd BW, et al. Use of a bronchoscopic protected catheter brush for the diagnosis of pulmonary infections. Chest 1982; 81:556. 73. Henriquez AH, Mendoza J, Gonzalez PC. Quantitative culture of bronchoalveolar lavage from patients with anaerobic lung abscesses. J Infect Dis 1991; 164:414. 74. Verma P. Laboratory diagnosis of anaerobic pleuropulmonary infections. Semin Respir Infect 2000; 15:114. 75. Sosenko A, Glassroth J. Fiberoptic bronchoscopy in the evaluation of lung abscesses. Chest 1985; 87:489. 76. Finegold FM, George WL, Mulligan ME. Anaerobic infections. Dis Mon 1985; 31:8. 77. Civen R, Jousimies-Somer H, Marina M, et al. A retrospective review of cases of anaerobic empyema and update of bacteriology. Clin Infect Dis 1995; 20:S224. 78. Wollenman OJ, Finland M. Pathology of staphylococcal pneumonia complicating clinical influenza. Am J Pathol 1943; 19:23. 79. Fisher Am, Trever RW, Curtin JA, et al. Staphylococcal pneumonia: a review of 21 cases in adults. N Engl J Med 1958; 258:919.

218

Bartlett

80. Francis JS, Doherty MC, Lopatin U, et al. Severe community-onset pneumonia in healthy adults caused by methicillin-resistant Staphylococcus aureus carrying the Panton-Valentine Leukocidin genes. Clin Infect Dis 2005; 40:1376–1378. 81. Micek ST, Dunne M, Kollef MH. Pleuropulmonary complications of Panton-Valentine leukocidin-positive community-acquired methicillin-resistant Staphylococcus aureus: importance of treatment with antimicrobials inhibiting exotoxin production. Chest 2005; 128:2732–2738. 82. Gillet Y, Issartel B, Vanhems P, et al. Association between Staphylococcus aureus strains carrying gene for Panton-vantine Leukocidin and highly lethal necrotizing pneumonia in young immunocompetent patients. Lancet 2002; 359:753–762. 83. Gonzalez BE, Hulten KG, Dishop MK, et al. Pulmonary manifestations in children with invasive community-acquired Staphylococcus aureus infection. Clin Infect Dis 2005; 41:583–590. 84. Lina G, Piemont Y, Godail-Gamot F, et al. Involvement of Panton-Valentine Leukocidinproducing Staphylococcus aureus in primary skin infections and pneumonia. Clin Infect Dis 1999; 29:1128–1132. 85. Finck-Barbancon V, Duportail G, Meunier O, Colin DA. Pore formation by a twocomponent leukocidin from Staphylococcus aureus within the membrane of human polymorphonuclear leukocytes. Biochim Biophys Acta 1993; 1182:275–282. 86. Bullowa JGM, Chess J, Friedman NJ. Pneumonia due to Bacillus friedlanderi. Arch Intern Med 1937; 60:735. 87. Wang JL, Chen KY, Fang CT, Hseuh PR, Yang PC, Chang SC. Changing bacteriology of adult community-acquired lung abscess in Taiwan: Klebsiella pneumoniae versus anaerobes. Clin Infect Dis 2005; 40:915–922. 88. Stack WA, Richardson PD, Logan RO, et al. Nocardia asteroids lung abscess in acute ulcerative colitis treated with cyclosporine. Am J Gastroenterol 2001; 96:225. 89. vanBurik JA, Hackman RC, Nadeem SO, et al. Nocardiosis after bone marrow transplantation: a retrospective study. Clin Infect Dis 1997; 24:1154. 90. Ernst A, Gordon FD, Hayek J, et al. Lung abscess complicates Legionella micdadei pneumonia in a liver transplant recipient: a case report and review. Transplantation 1998; 65:130. 91. Johnson KM, Huseby JS. Lung abscess caused by Legionella micdadei. Chest 1997; 111:252. 92. Senecal JL, St-Antoine R, Beliveau C. Legionella pneumophilia lung abscess in a patient with systemic lupus erythematosus. Am J Med Sci 1987; 293:309. 93. Kobayashi K. Streptococcus milleri as a cause of pulmonary abscess. Acta Paediatr 2001; 90:233. 94. Jerng JS, Hsueh PR, Teng LJ, et al. Empyema thoracic and lung abscess caused by viridians streptococci. Am Respir Crit Care 1997; 156:1508. 95. Frieden TR, Biebuyck J, Hierholzer WJ Jr. Lung abscess with group beta-hemolytic streptococcus. Case report and review. Arch Intern Med 1991; 151:1655. 96. Roy S, Kaplan EL, Rodriguez B, et al. A family of five cases of group A streptococcal pneumonia. Pediatrics 2003; 112:e61–e65. 97. Harvey RI, Sunstrum JC. Rhodococcus equi infection in patients with and without human immunodeficiency virus infection. Rev Infect Dis 1991; 13:139. 98. Verville TD, Huycke MM, Greenfield RA, et al. Rhodococcus equi infections of humans. 12 cases and a review of the literature. Medicine (Baltimore) 1994; 73:119. 99. Shapiro JM, Romney BM, Weiden MD, et al. Rhodococcus equi endobronchial mass with lung abscess in a patient with AIDS. Thorax 1992; 47:62. 100. Torres-Tortosa M, Arrizabalaga J, Villaneuva JL, et al. Prognosis and clinical evaluation of infection caused by Rhodococcus equi in HIV-infected patients: a multicenter study of 67 cases. Chest 2003; 123:1970–1976. 101. Chirinos JA, Lichtstein DM, Garcia J, Tamariz LJ. The evolution of Lemierre syndrome: report of 2 cases and review of the literature. Medicine (Baltimore) 2002; 81:458–465.

Pleural Empyema/Lung Abscess in the Critical Care Unit

219

102. Lemierre A. On certain septicaemias due to anaerobic organisms. Lancet 1936; 1:701. 103. Cook RJ, Ashton RW, Aughenbaugh GL, Ryu JH. Septic pulmonary embolism: presenting features and clinical course of 14 patients. Chest 2005; 128:162–166. 104. Veld DH, Wuthiekanun V, Cheng AC, et al. The role and significance of sputum cultures in the diagnosis of melioidosis. Am J Trop Med Hyg 2005; 73:657–661. 105. Aaron SD, Vandemheen KL, Ferris W, et al. Combinations antibiotic susceptibility testing to treat exacerbations of cystic fibrosis associated with multiresistant bacteria: a randomized, double-blind, controlled clinical trial. Lancet 2005; 366:433–435. 106. Canton R, Cobos N, de Gracia J, et al., On Behalf of the Spanish Consensus Group for Antimicrobial Therapy in the Cystic Fibrosis Patient. Antimicrobial therapy for pulmonary pathogenic colonization and infection by Pseudomonas aeruginosa in cystic fibrosis patients. Clin Microbiol Infect 2005; 11:690–703. 107. Levision ME, Mangura CT, Lober B, et al. Clindamycin compared with penicillin for the treatment of anaerobic lung abscess. Ann Intern Med 1983; 98:466. 108. Gudiol F, Manressa F, Pallares R, et al. Clindamycin vs. penicillin for anaerobic lung infections. Arch Intern Med 1990; 158:2525. 109. Germaud P, Poirier J, Jacqueme P, et al. Monotherapy using amoxicillin/clavulanic acid as treatment of first choice in community-acquired lung abscess during chest physical therapy: a case report. Phys Ther 1988; 68:371. 110. Allewelt M, Schuler P, Bolcskei PL, Mauch H, Lode H. Study Group on Aspiration Pneumonia. Clin Microbiol Infect 2004; 10:163–170. 111. Eykyn SJ. The therapeutic use of metronidazole in anaerobic infection: six years’ experience in a London hospital. Surgery 1983; 93:209. 112. Perlino CA. Metronidazole vs. clindamycin treatment of anaerobic pulmonary infection. Arch Intern Med 1981; 141:1424. 113. Sanders CV, Hanna BJ, Lewis AC. Metronidazole in the treatment of anaerobic infections. Am Rev Respir Dis 1979; 120:337. 114. Smith DT. Medical treatment of acute and chronic pulmonary abscesses. J Thorac Surg 1948; 17:72. 115. Cordice JW Jr., Chitkara RK. The role of surgery in treating pleuropulmonary suppurative disease—review of 77 cases managed at Queens Hospital Center between 1986 and 1989. J Natl Med Assoc 1992; 84:145. 116. Pfitzner J, Peacock JM, Tsirgiotis E, et al. Lobectomy for cavitating lung abscess with haemoptysis: strategy for protecting the contralateral lung and also the non-involved lobe of the ipsilateral lung. Br J Anaesth 2000; 85:791. 117. Tseng YL, Wu MH, Lin MY, et al. Surgery for lung abscess in immunocompetent and immunocompromised children. J Pediatr Surg 2001; 36:470.

11 Infective Endocarditis and Its Mimics in the Critical Care Unit John L. Brusch Department of Medicine, Harvard Medical School, Cambridge, Massachusetts, U.S.A.

OVERVIEW Since Osler’s comprehensive description of infective endocarditis (IE) in the 1880s, this disease has continuously evolved in respect to its epidemiology, clinical presentations, and therapy. Over the last 30 years, greater numbers of patients with IE are being cared for in critical care units (CrCU) mainly because of the increased incidence of acute staphylococcal IE. In recent series, approximately 60% of cases of IE are caused by Staphylococcus aureus (1). By prolonging the lives of those with acute IE, antibiotics are contributing to the increasing number of cardiac and extracardiac complications of this type of valvular infection. Even in subacute IE, antibiotics have failed to lessen the frequency of embolic complications including mycotic aneurysms (2). This is due to the delay in diagnosis that has not lessened over the last 30 years. The average interval between the onset of valvular infections and diagnosis remains six weeks (3). Although complications of IE affect the heart most frequently, neurological events and sepsis are the most frequent causes of death. Increasingly, patients being cared for in CrCU are developing IE that is a consequence of the increased reliance on various types of intravascular devices in the care of the critically ill. Even the surgical procedures meant to repair the damage of cardiac infections pose their own unique risks to the patient. This chapter will focus on the epidemiology, pathogenesis presentation, diagnosis, treatment prevention of the bacteria, and the types of IE that will most likely be encountered in CrCU. Among these are native valve IE (NVE), prosthetic valve endocarditis (PVE), and health-care associated IE (HCIE). The organisms that will be discussed include the various streptococcal species: Streptococcus viridans, the nutritionally variant streptococci (NVS) and group B streptococci, gram-negative aerobes, and fungi and of course S. aureus. Many infectious and noninfectious disease processes share important clinical features with IE. The most effective of these mimics will also be highlighted.

221

222

Brusch

Epidemiology IE is usually an infection of the valvular endocardium, rarely of the mural endocardium. Because it can present with noncardiac signs and symptoms, especially subacute disease, the diagnosis of IE may be particularly challenging. Additionally 5% to 10% of cases may be blood culture negative. Histological exam of the involved valvular tissue remains the gold standard for diagnosing IE (4). Because it is uncommon to obtain premortem valvular tissue, Von Reyn et al. developed strict case definitions for the diagnosis of IE (5). By these criteria, only 20% of clinically diagnosed cases could be considered as definite examples of IE. In order to increase this system’s sensitivity, The Duke Endocarditis Service in 1994 combined echocardiographic findings with a set of clinical measures (6). These Duke Criteria have far greater positive predictive value and negative predictive value of at least 92% (6). These criteria were further modified in 2000 (see Section on diagnosis). The incidence of IE has not changed over the last 50 years (approximately 4/ 100,000 person-years) (7). Those institutions that serve large numbers of intravenous drug abusers (IVDA) and those that perform large numbers of intravascular procedures will generate many more cases than will a community hospital. The incidence of IE is at least two times more common in men than in women. In those greater than 50 years of age, the male incidence is six times that of the female (1,7,8). The risk of developing HCIE is equal among men and women (9). The term HCIE is preferable to that of nosocomial IE because it recognizes the fact that a growing amount of medical care, such as hemodialysis, is provided outside of the hospital proper. IE has become a disease of the older population. In a series collected in the 1990s (1), the mean age was 50 with 35% greater than 60 years of age and 15% greater than 70 years of age. The major exception to this ‘‘graying’’ trend is IE among IVDA of which 85% are complicated with HIV infection. Several factors have contributed to this rise of IE among elderly (10,11). Among these are: 1. The aging-related atherosclerotic changes to the circulatory system (i.e., calcific valvular disease). 2. The proliferation of cardiac surgery and placement of intravascular devices in older individuals (the majority of patients with hospital acquired staphylococcal bacteremia are in this age group). 3. Those with congenital heart disease are living longer. 4. The age associated dysfunction of the immune system. There has been a marked increase in cases of HCIE, IVDA IE, and PVE accounting for 22%, 36%, and 16%, respectively, of all cases (1). Perhaps the most striking change is the rise in cases of acute IE. Currently these account to for more than 50% of cases and growing (1). This reflects the rise of staphylococcal/health-care associated bloodstream infections (HCBSI) (12) coupled with a significant decrease in disease caused by S. viridans (13). Predisposing Cardiac Lesions Acquired and Congenital Cardiac Abnormalities The frequency of underlying cardiac lesions in IE of native tissue is dependent on the prevalence of acute IE in the studied population. More than 50% of cases of acute IE have no definable underlying cardiac pathology (14). Congenital heart disease is found in 15% of cases. Congenital bicuspid aortic valve is the most common example

Infective Endocarditis and Its Mimics in the Critical Care Unit

223

found in older patients (20% of cases) (15). A somewhat neglected predisposing condition is asymmetric septal hypertrophy. This accounts for 5% of cases of IE (16). The greater the degree of obstruction the greater is the chance of valvular infection. The mitral valve is the most frequently affected (17). Other contributory congenital conditions are ventricular septal defect, patent ductus arteriosus, and tetralogy of Fallot (18). In the developed world, rheumatic heart disease (RHD) accounts for 60 years of age (21). These lesions include calcified valvular leaflets, calcified mitral rings, and postmyocardial infarction thrombi. Calcified aortic stenosis may result from calcium deposition either on a congenital bicuspid valve or on a previously normal valve damaged by hemodynamic stress over the years (22). Because of the prevalence of associated illnesses, such as diabetes or renal failure, IE in this group of patients has a poorer than average outcome. The degree of stenosis is often not hemodynamically significant and so is frequently overlooked as a candidate for antibiotic prophylaxis (23). Forty percent of NVE, excluding IVDA IE, infect only the mitral valve and 40% the aortic solely. The right side of a heart is rarely involved except in cases of IVDA IE (24). Overall, PVE accounts for 10% to 20% of all cases of IE and 26% in those >60 years of age (25). The 10-year risk of infection of both mechanical and bioprosthetic valves is approximately 5% (26). During the first three months after implantation, mechanical valves are more at risk. However, after one year, bioprosthetic valves exhibit a greater chance of IE due to the ongoing calcification of the leaflets associated with degeneration of the valvular tissue. Infection of implanted pacemakers (PMs) and cardioverter-defibrillator is becoming more frequent (27). These devices most often become infected within a few months of placement. Infection of PMs may involve the generator pocket (the most common type): there may be infection of proximate leads and those portions of the leads that are in direct contact with the endocardium. The last type is synonymous with true Pacemaker IE (PMIE) (0.5% of all PMs) (28). Seventy five percent of all types of PM infections are caused by staphylococci. It is important to note that the true risk for development of IE for most underlying cardiac abnormalities has not been accurately quantified for many cardiac conditions (29). Patients with MVP and associated regurgitation have a three- to eightfold increase in risk. Those with a history of prior IE or with prosthetic valves in place have a 60- to 185-fold increased risk. Extracardiac Predisposing Factors The incidence of IE among IVDA lies between 1% and 5% per year. IVDA IE causes up to 20% of hospital admissions and 5% to 10% of deaths among narcotics abusers. It appears that IVDA IE is decreasing due to safer injection practices, such as the use of sterile needles and syringes brought about to combat the spread of the HIV virus among this population. Eighty percent of patients are male, probably due to their

224

Brusch

longer usage of illicit drugs. The tricuspid valve is infected in about 70% of cases with the mitral and aortic valves involved in 20% to 30% of patients. The pulmonic valve is rarely infected. In some series, up to 75% of IVDA IE occurs in the absence of any preexisting valvular abnormalities. Forty three percent of recurrent valvular infections are seen in illicit drug users. Interestingly 7% had prosthetic valves in place. A history of IVDA IE has become the most significant risk factor for recurrent NVE (30–35). HCIE has been defined as a valvular infection that presents either 48 hours after the patient has been hospitalized or that is associated with a health-care facilitybased procedure that has been performed within four weeks of presentation (36,37). Patients with HCIE are older and have a higher rate of underlying valvular disorders and develop bacteremias that are related to a variety of invasive vascular procedures. The incidence of HCIE accounts for approximately 20% of IE overall (1,12) and appears to be ever increasing. Much of this is due to the rise in staphylococcal bacteremia associated with intravascular line infection (12,38,39). HCIE often involves normal valves. Type 1 HCIE results from endocardial trauma to the right ventricular wall produced by an intravascular catheter. Type 2 HCIE occurs in patients who develop left-sided IE due to bacteremias originating from the skin, urinary tract or intravenous lines, or other invasive procedures. In this situation, the left side predominates because of the greater proportion of structural abnormalities on that side (i.e., degenerative valvular disease, MVP). In addition to S. aureus and coagulasenegative staphylococci (CONS), gram-negative organisms and fungi are frequently involved. HCIE may be fatal in up to 50% of cases as compared to an overall mortality rate of 11% of IE acquired in the community. This increase in mortality is attributable in part to the older age of the patients with HCIE (64% of patients >60 years of age) (40). An exception to this is that community-acquired cases of S. aureus IE have a higher mortality rate than S. aureus HCIE. This may be due in part to the higher rate of metastatic complications arising from the prolonged bacteremia prior to the correct diagnosis being made (41). The major reason for focusing on HCIE has been well expressed by Friedland et al. (9), ‘‘nosocomial endocarditis occurs in a definable subpopulation of hospitalized patients and is potentially preventable.’’ Various types of primary immunodeficiencies as well as diseases that lower the patient’s resistance to infection by a variety of mechanisms have been cited as predisposing risk factors for developing IE (42). Among them are a variety of neoplasms, diabetes mellitus, renal failure, liver disease, and the use of corticosteroids. All of these disease states are associated with an increased frequency of bacteremias (43). The role of dental procedures as a risk factor for developing IE is currently in question (44) and will not be further discussed, especially because most of the patients who develop the IE due to bacteremias secondary to dental work do not require care in CrCU. Suffice it to say that the vast majority of dental associated IE are secondary to an increase in transit bacteremias that arise from poor dental hygiene and are not due to any one specific procedure. Table 1 presents the recent changes in the characteristics of IE.

MICROBIOLOGY The exact profile of causative organisms of a given hospital is dependent on the population it serves. The pathogens causing NVE are somewhat different than those that produce PVE or IVDA IE (Table 2). Overall, S. aureus produces about 30%

Infective Endocarditis and Its Mimics in the Critical Care Unit

225

Table 1 Changing Patterns of Infective Endocarditis Since 1966 Marked increase in the incidence of acute IE Rise of nosocomial, IVDA and prosthetic valve IE Change in the underlying valvular pathology: RHD 60 yrs of age) The increased numbers of vascular procedures Abbreviations: IE, infective endocarditis; IVDA, intravenous drug abusers; RHD, rheumatic heart disease.

of cases, CONS 16%, S. viridans 23%, S. bovis 5%, Enterococcus faecalis 4%, gramnegative organisms 2%, no growth 90% of cases (154). An increasingly common problem in CrCU is the management of S. aureus bacteremia in the presence of an intravascular catheter. Approximately 25% of these cases represent IE. Separating S. aureus IE from cases of uncomplicated staphylococcal bacteremia is essential for determining the length of therapy after removal of short-term lines and determining whether long-term lines need to be removed at all. Hematuria, associated with S. aureus bacteriuria, is a useful indicator of sustained S. aureus bacteremia. Hematuria is the result of embolic renal infarction or immunologically mediated glomerulonephritis (155). The presence of intracellular bacteria on smears of blood drawn through intravascular catheters is specific for infected devices (156). TEE provides the most specific means of distinguishing uncomplicated S. aureus bacteremia from valvular infection. Twenty-three percent of catheter associated staphylococcal bacteremia have evidence of IE on TEE even in the absence of clinical or positive TTE findings. Table 11 presents an approach to management of short-term intravascular catheter associated S. aureus bacteremia (157). Persistent bacteremia after three days of appropriate antibiotic therapy is an independent risk factor for endocarditis as well as death (158). Surgically implanted long-term catheters (Broviac or Hickman) do not need to be automatically removed except in the presence of IE, infection of the vascular tunnel, suppurative thrombophlebitis or pathogens such as Corynebacterium JK, Pseudomonas species, fungi, S. aureus, or mycobacteria (159). Intraluminal infusions of appropriate antibiotics have at least 30% to 50% greater success against sensitive organisms. The use of thrombolytic agents to dissolve the fibrin sheath of the catheter appears to improve the efficacy of the infused antibiotic (137).

Antibiotic Therapy There are many more challenges to sterilizing an infected thrombus with antibiotic therapy than to sterilizing a large infected vegetation. Among these factors are (i) the density of organisms (10–100 billion bacteria/g of tissue) and (ii) the decreased metabolic and replicative activity of the intrathrombus organisms that make the bacteria less sensitive to the action of most antibiotics (160). In addition, the mobility and phagocytic function of white cells is impaired in the fibrin-rich vegetation.

246

Brusch

Table 11 Management of Staphylococcus aureus Bacteremia in the Presence of an Intravascular Catheter Prompt removal of the catheter Institution of appropriate antibiotic therapy Follow-up blood cultures within 24–48 hrs If follow-up blood cultures are negative and The TEE shows no signs of IE There is no evidence of metastatic infection Then 2 wks of antibiotic therapy would be appropriate If follow-up blood cultures are positive and The TEE shows signs of IE Then 4 wks of intravenous therapy is appropriate If follow-up blood cultures are positive and The TEE shows no signs of IE Further imaging studies should be performed to rule out other sources of bacteremia (osteomyelitis, mediastinitis, splenic abscess) Abbreviations: TEE, transesophageal echocardiography; IE, infective endocarditis.

The basic principles of antibiotic therapy of IE include: 1. The necessity to employ bactericidal antibiotics because of the ‘‘hostile’’ environment of the thrombus. 2. The minimal inhibitory concentration (MIC) and the minimal bacteriostatic concentration (MBC) of the pathogen need to be determined to insure adequate overkill. It is estimated that in the case of Escherichia coli IE, 220 times the MBC of ceftriaxone is required to sterilize the vegetation (161). Determining the bactericidal titer should be limited to those patients who are not responding well to therapy or who are infected by an unusual organism. 3. In general, intermittent dosing of an antibiotic provides superior penetration of tissue as compared to continuous infusion. Its penetration into tissue is directly related to its peak level in serum (162). 4. All patients should be initially treated in a health-care facility for one to two weeks to monitor for hemodynamic stability. 5. In the case of acute IE, antibiotic therapy should be started after three to five sets of blood cultures are drawn within 60 to 90 minutes so as to minimize valvular damage. The selection of the antibiotic regimen is based on the clinical history and physical examination. 6. For cases of subacute IE, treatment may be delayed until the final culture and sensitivity results are available because a delay of one to two weeks does not adversely affect the final outcome. 7. Usually duration of therapy ranges from four to six weeks. A four-week course is quite appropriate for uncomplicated NVE (for sensitive S. viridans this sometimes can be decreased to two weeks—see below). Six weeks are required for the treatment of PVE and in those infections with large vegetations such as caused by the HACEK family (117). A daily temperature maximum of >37 C for 10 days into treatment merits concern. This situation may represent a relatively resistant pathogen, extracardiac infection, pulmonary or systemic emboli, drug fever, Clostridium difficile colitis or

Infective Endocarditis and Its Mimics in the Critical Care Unit

247

an infected intravenous site (163). If the pathogen is not sensitive to the administered antibiotic, a thorough search for the cause should be conducted. Mycotic aneurysms are the most difficult to detect. A TTE should be performed. If that is not helpful, then a TEE should be performed (164). Complications, produced by embolic immunological events, are not necessarily related to the failure of treatment for valvular infection itself (165). Relapse of IE usually occurs within two months of cessation of treatment (166). The risk of relapse is greatly dependent on the infecting organisms. Appropriately treated NVE, caused by S. viridans, rarely relapses. Four percent of S. aureus IE and 30% of enterococcal IE relapse. Gram-negative organisms, especially P. aeruginosa, have higher rates of relapse (167). IE of >3 months duration prior to antibiotic treatment also has a high rate of relapse. The greatest risk factor for recurrence of IE is a prior case of IE (168). The second most common factor is a past history of IVDA IE. Forty percent of cases of IVDA IE are recurrent. Isolates of S. viridans classically have been quite sensitive to the beta-lactam antibiotics, the aminoglycosides (gentamicin and streptomycin), and vancomycin. Valvular infection caused by these organisms may be cured by a two-week course of a beta-lactam antibiotic combined with gentamicin (169). To undertake a short course approach, the following conditions must exist: a sensitive S. viridans (MIC < 0.1 mcg/ml), NVE of < 3 months duration, vegetation size < 10 mm in diameter, no cardiac or extracardiac complications, a low risk for developing aminoglycoside nephrotoxicity, and good clinical response within the first week of therapy. There is a growing amount of S. viridans isolates that are resistant to penicillin (MIC > 0.1 mg/mL). Highly resistant isolates have a MIC that is >1 mg/mL. Thirteen percent of S. viridans isolates in this country are highly resistant to penicillin, with 70% being resistant to ceftriaxone (170). All NVS are resistant to penicillin, many being highly resistant. Many penicillinsensitive strains are tolerant to the beta-lactam drugs. Tolerance is a phenomenon in which the MBC of antibiotic exceeds its MIC by a factor of 10 (171). Against these isolates, the penicillins behave practically as bacteriostatic compounds. Although penicillin by itself can cure most cases of S. viridans, the third generation cephalosporin, ceftriaxone because of its pharmacokinetics, is the antibiotic of choice because of its twice a day dosing regimen. The combined use of penicillin or glycopeptide with gentamicin is needed to eradicate resistant streptococci. Tolerant streptococci are best managed by a combination of penicillin and gentamicin. Table 12 summarizes the recommendations for treatment of nonenterococcal streptococci. Since the advent of antibiotics, enterococci have posed major resistance problems due to their ability to develop multiple resistance mechanisms. They are resistant to all the cephalosporins and to the penicillinase-resistant penicillins such as nafcillin and oxacillin. Penicillin and ampicillin are ineffective when used singly against serious enterococcal infection. Aminoglycosides because of their failure to penetrate the bacterial cell wall are ineffective when used alone (172). The success of the serendipitously recognized combination of penicillin and streptomycin opened up the whole field of synergy. The cell wall active antibiotic allows penetration of the aminoglycosides into the bacterial anterior and reach its target, the ribosome. A serum concentration of 3 mg/mL of gentamicin is necessary for synergism. If the isolate is resistant to ampicillin or penicillin synergism is not possible. Currently, 5% of E. faecalis and 40% of E. faecium possess high-grade resistance to gentamicin (>2000 mg/mL) (173). Resistance to streptomycin has been

248

Brusch

Table 12 Guidelines for Antimicrobial Therapy of Nonenterococcal Streptococcal Infective Endocarditisa Antibiotic Penicillin-sensitive Streptococcus viridans Penicillin Gb Penicillin Gb and gentamicinc

Or ceftriaxone Penicillin-resistant or -tolerant S. viridans Penicillin Gb Gentamicin NVS and group B streptococci Penicillin Gb Gentamicin

Dosage regimen Penicillin G 20,000,000 U IV in four divided doses for 4 wks Penicillin 20,000,000 U IV in four divided doses for 2 wks gentamicin 3 mg/kg given q24hrs as a single dose or in divided doses q8hrs for 2 wks (ceftriaxone 2 g IV/1 M for 4 wks may be used in patients with mild reactions to penicillin) Ceftriaxone 2 g IV/1 M for 4 wks (may be used in patients with mild reactions to penicillin) Penicillin G 20,000,000 U IV in four divided doses for 4 wks Gentamicin 3 mg/kg given q24hrs as a single dose or in divided doses q8hrs for 2 wks Penicillin G 20,000,000 U IV in four divided doses for 6 wks Gentamicin 3 mg/kg given q24hrs as a single dose or in divided doses q8hrs for 2 wks

Note: Drug dosages: a For patients with normal renal function. b Vancomycin 30 mg/kg IV q12hrs in patients highly allergic to penicillin. c Short course therapy (see text). Abbreviation: NVS, nutritionally variant streptococci.

prevalent for a long time. Some gentamicin-resistant isolates remain sensitive to streptomycin and vice versa. Ampicillin resistance, due to beta-lactamase production, has been recognized since the 1980s. This may not be detectable by routine sensitivity testing. In the absence of ampicillin/penicillin, vancomycin, and aminoglycoside resistance, the combination of a cell wall active antibiotic and an aminoglycoside remains the preferred therapeutic approach. In the setting of normal renal function, the daily dose of ampicillin is 4 g given IV every eight hours. Gentamicin (1.5 mg/kg) is to be given every 12 hours (174). Vancomycin (1 g IV every 12 hours) is substituted for ampicillin in those allergic to penicillin or for isolates resistant to ampicillin. When resistance to both gentamicin and streptomycin is present, continuously infused ampicillin, to achieve a serum level of 60 mg/mL, appears to be the best option (175). Quinupristin/dalfoprastin and linezolid should be considered as alternatives. They have the disadvantage of being bacteriostatic antibiotics against the enterococcus (176). In addition, Quinupristin/dalfoprastin is active only against E. faecium and not against E. faecalis, the most common enterococcal species. Daptomycin appears to be bactericidal against these organisms, but experience is quite limited in treating IE with this drug (177) (see table for dosages of these antibiotics).

Infective Endocarditis and Its Mimics in the Critical Care Unit

249

It is extremely important to emphasize the need to obtain MICs and MBCs for ampicillin, the aminoglycosides and vancomycin in all cases of enterococcal IE to arrive at the best therapeutic approach. Staphylococcus aureus The penicillinase-resistant penicillins are the drugs of choice in treating MSSA IE; vancomycin is significantly less effective than these compounds against MSSA. It should be used only in valvular infections caused by MRSA or for patients who

Table 13 Antibiotic Therapy of Staphylococcus aureus Infective Endocarditisa Valve type (IE type) Native (MSSA)

Antibiotic Oxacillin þ/ gentamicin

or Vancomycinb,c þ/ gentamicin

or Cefazolin þ/ gentamicin

Prosthetic (MSSA)

Oxacillin or vancomycin or cefazolin and

Rifampin and Gentamicin

Native (MRSA)

Vancomycinc

Prosthetic (MRSA)

Vancomycinc and Rifampin and Gentamicin

a

Dosage Oxacillin 2 g IV q4hrs for 4–6 wks þ/ gentamicin 2 mg/kg q24hrs as a single dose or in divided doses q8hrs for 5 days Vancomycin 15 mg/kg IV q12hrs for 4–6 wks  gentamicin 3 mg/kg q24hrs as a single dose or in divided doses q8hrs for 5 days Cefazolin 1.5 g IV q8hrs for 4–6 wks (in patients with mild allergies to penicillin)  gentamicin 3 mg/kg q24 hrs as a single dose or in divided doses q8hrs for 5 days Oxacillin 2 g IV q4hrs for 4–6 wks or vancomycin 15 mg/kg IV q12hrs for 4–6 wks or cefazolin 1.5g IV q8hrs for 4–6 wks in patients with mild allergies to penicillin Rifampin 300 mg PO q8hrs for 6 wks Gentamicin 3 mg/kg q24hrs as a single dose or in divided doses q8hrs for 2 wks Vancomycin 15 mg/kg IV q12hrs for 4–6 wks Vancomycin 15 mg/kg IV q12hrs for 4–6 wks Rifampin 300 mg PO q8hrs for 6 wks Gentamicin 3 mg/kg q24hrs as a single dose or in divided doses q8h for 2 wks

For patients with normal renal function. For patients with severe penicillin allergy. c Substitute linezolid in critically ill patients or those with significant renal failure (refer to discussion in text). Abbreviations: IE, infective endocarditis; MSSA, methicillin-sensitive S. aureus; MRSA, methicillinresistant S. aureus. b

250

Brusch

Table 14 Therapy for Coagulase-Negative Staphylococcal Prosthetic Valve Endocarditisa Antibiotic Vancomycin Rifampin Gentamicin

Dosage regimen 15 mg/kg q12hrs for 6 wks 300 mg PO q8hrs for 6 wks 3 mg/kg q24hrs IV as a single dose or in divided doses q12hrs for 2 wks

a

80% of isolates recovered within the first year after valve replacement are resistant to the beta-lactam antibiotics. After this period, 30% are resistant. Sensitivity to the penicillins must be confirmed because standard sensitivity testing may not detect resistance. If the isolate is sensitive, oxacillin or cefazolin may be substituted.

are significantly allergic to penicillin. Vancomycin has a failure rate of 35% in MSSA IE (178). Although cefazolin is used in treating MSSA IE, especially in patients with mild allergic reactions to penicillins, it should be administered with realization that there have been failures with this drug. This is probably due to production of type A beta-lactamase by the pathogen (179). For the first three to five days of treatment of MSSA or MRSA IE, the addition of gentamicin to the penicillin or vancomycin should be strongly considered in patients who are not at increased risk of aminoglycoside nephrotoxicity. This combination has not been proven to decrease overall mortality. In decreasing the duration of bacteremia and fever, it may minimize the intra- and extra-cardiac complications of S. aureus IE (180). Right-sided IVDA IE, caused by MSSA, has been successfully treated in two weeks of intravenous therapy with a combination of nafcillin/oxacillin and gentamicin (181). This may be due to the fact that in right-sided endocarditis, antibiotic penetration of the vegetation is greater, and there is a lower concentration of bacteria than in left-sided disease due to the lower oxygen tension in the right ventricle. Those cases of IVDA IE in which the patient is HIV positive or there is evidence of left-sided disease or of lung abscess or of other metastatic sites of infection are not suitable for a short course of antibiotic therapy. To achieve best outcomes for staphylococcal PVE due to MSSA, MRSA, or CONS, a triple drug approach is advised. Rifampin is the key component. Rifampin has the distinctive ability to kill staphylococci adherent to prosthetic material as well as being able to penetrate phagocytes (182). The other two agents are chosen because of their activity against the target isolate with the aim of preventing the development Table 15 Antibiotic Treatment Options for Treatment of Endocarditis Due to Highly Resistant Gram-Positive Organismsa Antibiotic dosage Linezolid 600 mg q24hrs (either PO or IV)b Quinpristin/dalfopristin 7.5 mg/kg q8hrs Daptomycin 6 mg/kg q24hrsc a

See text for indications. Effectiveness of the PO route may approximate that of the IV route (see text). c This higher dose (usual dose ¼ 4 mg/kg q24h) is probably required to treat infective endocarditis due to Staphylococcus aureus and enterococci. Source: From Ref. 188. b

Infective Endocarditis and Its Mimics in the Critical Care Unit

251

Table 16 Therapy of Various Types of Infective Endocarditisa Organism Culture negative

Pseudomonas aeruginosa

HACEK group

Antibiotic regimen Ampicillin 2 g IV q4hrs for 4 wksb Gentamicin 5 mg/kg q24hrs IV given in a single dose or in divided doses q8hrs for the first 2 wks Oxacillin 2 g IV q4hrs for 4 wks Or if MRSA is suspected or prosthetic material is present, vancomycin 30 mg/kg q 12hrs for 4 wks Ticarcillin 3 g IV q4hrs for 6 wksb Tobramycin 5 mg/kg q24hrs IV given in a single dose or in divided doses q8hrs Ampicillin 2 g IV q4hrs for 4–6 wksb Gentamicin 5 mg/kg q24hrs as a single dose or in divided doses q8hrs

Alternative regimen Culture negative

Ceftazidimec 2 g IV q8hrs for 6 wks Aztreonamd 2 g IV q6hrs for 6 wks Tobramycin 5 mg/kg IV q24hrs given in a single dose or in divided doses q8hrs Cefotaximec 2 g IV q8hrs for 4–6 wks Gentamicin 5 mg/kg q24hrs given in a single dose or in divided doses

a

For patients with normal renal function. Preferred regimen (see text). c In patients with mild penicillin allergy. d In patients with severe penicillin allergy. Abbreviations: MRSA, methicillin-resistant S. aureus; HACEK, Haemophilus, A. actinomycetemcomitans, C. hominis, and Kingella. b

Figure 1 Approach to the patient at risk for candidal endocarditis. Source: From Ref. 192.

252

Brusch

Table 17 Recommendations for the Prevention of Intravascular Catheter-Related Infections General Not recommended Preventive strategies incorporating therapeutic antimicrobial agents During catheter insertion Strongly recommended Full-barrier precautions during central venous catheter insertion Subcutaneous tunneling short-term catheters inserted in the internal jugular or femoral veins when catheters are not used for blood drawing Contamination shield for pulmonary artery catheters. Insertion-site preparation with chlorhexidine-containing antiseptics Prophylaxis with vancomycin and other therapeutic agents Recommended Subclavian vein, rather than jugular or femoral vein, catheter insertion Consider Insertion-site preparation with tincture of iodine Full-barrier precautions during insertion of midline, peripheral artery, and pulmonary artery catheters Not recommended Femoral vein catheter insertion Catheter maintenance Strongly recommended Provide-iodine ointment applied to hemodialysis catheter-insertion sites Specialized nursing teams caring for short-term peripheral venous catheters at institutions with a high incidence of infection Chlorhexidine-silver sulfadiazine-impregnated short-term central venous catheters Minocycline-rifampin-impregnated short-term central venous catheters Antiseptic chamber-filled hub or hub-protective antiseptic sponge for central venous catheters with an expected duration of approximately 2 wk Povidone-iodine-saturated sponge enclosed in plastic casing fitted around the central venous catheter hubs Assess need for intravascular catheters on a daily basis; remove catheters as soon as possible after intended use Adequate nurse-to-patient ratio in ICUs Change needleless system, the device and endcap if present on a regular basis in accordance with manufacturers’ guidelines and reduce contact with nonsterile water Continuing quality-improvement programs to improve compliance with catheter care guidelines Disinfect catheter hubs and sampling ports before accessing Low-dose heparin for patients with short-term central venous catheters Low-dose warfarin for patients with long-term central venous catheters Pulmonary artery catheters heparin-bonded with benzalkonium chloride. Povidone-iodine ointment applied to nontunneled, long-term central venous or midline catheter-insertion sites of immunocompromised patients with heavy Staphylococcus aureus carriage (i.e., patients with AIDS and cirrhosis) Specialized nursing teams caring for catheters used for Total parenteral nutrition (TPN) Recommended Gauze dressings preferred if excessive oozing of blood from insertion site Consider Antiseptic chamber-filled hub or hub-protective antiseptic sponge for central venous catheters in ICUs (Continued)

Infective Endocarditis and Its Mimics in the Critical Care Unit

253

Table 17 Recommendations for the Prevention of Intravascular Catheter-Related Infections (Continued ) Not recommended Routine replacement of central venous catheters Mupirocin ointment applied to the insertion site Triple antibiotic ointments applied to the insertion sites Silver-impregnated, subcutaneous collagen cuffed central venous catheters Inline filters for prevention of catheter infection Abbreviation: TPN, total parenteral nutrition. Source: From Ref. 194.

of rifampin-resistant organisms. For staphylococci resistant to gentamicin, a fluoroquinolone may be effective (183). Linezolid appears to have superior outcomes to vancomycin for many types of MRSA infections (184). Several studies indicate that this is the case for MRSA IE (185,186). In seriously ill patients, strong consideration should be given to substituting linezolid for vancomycin. Some studies indicate that there is an advantage to combining linezolid with gentamicin or imipenem. However in average doses linezolid is a bacteriostatic drug with case reports of failure to cure MRSA IE (187). Because of its bactericidal properties, daptomycin is quite promising for the treatment of MRSA IE (188). More experience must be gained with this antibiotic and special attention paid to the myositis associated with its use. Tables 13 and 14 summarize the antibiotic treatment of MSSA, MRSA, and CONS infections of both native and prosthetic valves. Table 15 presents therapeutic options for treating highly resistant gram-positive organisms. Table 16 summarizes the antibiotic approach to other types of IE that may be encountered in the CrCU. Fungal Endocarditis Combined medical and surgical therapy is necessary for cure of most cases of fungal IE. Amphotericin B has been the mainstay of medical therapy (189). However the newer antifungals, caspofungin and voriconazole, hold promise as less toxic and more effective alternatives to the older compound (190,191). Figure 1 presents the approach to the patient at risk for candidal endocarditis. Prophylaxis of IE in the CrCU Prophylaxis of CrCU IE should be focused on limiting the rate of line related bacteremia in addition to the more traditional methods (193). Table 17 summarizes the CDC’s recommendations to prevent this type of infection (194).

REFERENCES 1. Bouza E, Menasalvas A, Munoz P, et al. Infective endocarditis–A prospective study at the end of the twentieth century. Medicine 2001; 80:298–307. 2. Mansur A, Grinberg M, Da Luz P, et al. The complications of infective endocarditis: a reappraisal in the 1980s. Arch Intern Med 1992; 152:2428–2432. 3. Starkebaum M, Durack D, Beeson P. The ‘‘incubation period’’ of subacute bacterial endocarditis. Yale J Biol Med 1977; 50:49–60.

254

Brusch

4. Cunha BA, Gill V, lazar IM. Acute infective endocarditis. Diagnostic and therapeutic approach. Infect Dis Emerg 1996; 10:811–834. 5. Von Reyn CF, Levy BS, Arbeit RD, et al. Infective endocarditis: an analysis based on strict case definitions. Ann Intern Med 1981; 94(Part 1):505–518. 6. Durack DT, Lukes BS, Bright DK. Duke Endocarditis Service. New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Am J Med 1994; 96:200–209. 7. Berlin JA, Abrutyn E, Strom BL, et al. Incidence of infective endocarditis in the Delaware Valley, 1988–1990. Am J Cardiol 1995; 76(9):33–36. 8. King H, Harkness JL. Infective endocarditis in the 1980s. Part 1. Aetiology and diagnosis. Med J Aust 1984; 144:536–540. 9. Friedland G, Von Reyn CF, Levy BS, et al. Nosocomial endocarditis. Infect Control 1984; 5:284–288. 10. Terpenning MS, Buggy BP, Kauffman CA. Infective endocarditis: Clinical features in young and elderly patients. Am J Med 1987; 83:626–634. 11. Gladstone JL, Rocco R. Host factor and infectious diseases in the elderly. Med Clinic North Am 1976; 60:1225–1246. 12. Gaynes R. Health care-associated bloodstream infections: a change in thinking. Editorial. Ann Intern Med 2003; 137:850–851. 13. Cabell CH, Jollis JG, Peterson GE, et al. Changing patient characteristics and the effect on mortality in endocarditis. Arch Intern Med 2002; 162:90–95. 14. Kaye D. Changing patterns of infective endocarditis. Am J Med 1985; 78(suppl 6b): 157–162. 15. Delahaye JP, Loire R, Milton H, et al. Infective endocarditis on stenotic aortic valves. Eur Heart J 1988; 9(suppl E):S43–S49. 16. Chagnac A, Rudinki C, Loebel H, et al. Infectious endocarditis in idiopathic hypertrophic subaortic stenosis: report of three cases and review of the literature. Chest 1982; 81:346–361. 17. Hickey AJ, McMahon SW, Wilcken D. Arch without prolapse and bacterial endocarditis: when is antibiotic prophylaxis necessary? Am Heart J 1985; 109:431–435. 18. McKinsey DS, Ratts TE, Bisno Al. Underlying cardiac lesions in adults with infective endocarditis. The changing spectrum. Am J Med 1987; 82:681–688. 19. Starke J. Infections of the heart: infective endocarditis in children. In: Feigin R, Cherry J, eds. Pediatric Infectious Disease. 3rd ed. Philadelphia: WB Saunders, 1992:326–346. 20. Zuppiroli A, Rinaldi M, Kramer-Fox R, et al. Natural history of mitral valve prolapse. Am J Med 1995; 75:1028–1032. 21. Sandre RM, Shafran SD. Infective endocarditis: review of 135 cases over 9 years. Clin Infect Dis 1996; 22:276–286. 22. Selzer A. Changing aspects of the natural history of valvular aortic stenosis. N Engl J Med 1987; 217:91–98. 23. Bansal RC. Infective endocarditis. Medicine 1995; 79:1205–1246. 24. Nager F. Changing clinical spectrum of infective endocarditis. In: Horstkotte D, Bodnar E, eds. Infective Endocarditis. London: ICR Publishers, 1991:25–29. 25. Rutledge RR, Leim BJ, Applebaumb RE. Actuarial analysis of the risk of prosthetic valve endocarditis in 1598 patients with mechanical and bioprosthetic valves. Arch Surg 1985; 120:469–472. 26. Calderwood SB, Swinski LA, Waternaux CM, et al. Risk factors for the development of prosthetic valve endocarditis. Circulation 1985; 72:31–37. 27. Eggimann P, Waldvogel FA. Pacemaker and defibrillator infections. In: Waldvogel FA, Bisno AL, eds. Infections Associated with Indwelling Medical Devices. Washington, D.C.: American Society for Microbiology Press, 2000:247–264. 28. Duval X, Selton-Suty C, Alla F, et al. Endocarditis in patients with a permanent pacemakers: a 1-year epidemiological survey of infective endocarditis due to valvular and/or pacemaker infection. Clin Infect Dis 2004; 39:68.

Infective Endocarditis and Its Mimics in the Critical Care Unit

255

29. Steckelberg JM, Melton LJ III, Ilstrup DM, et al. Influence of referral bias on the apparent clinical spectrum of infective endocarditis. Am J Med 1990; 88:582–588. 30. Spijerkerman IJ, van Ameidjen EJ, Mientjes GH, et al. Human immunodeficiency virus infection and other risk factors for skin apices and endocarditis among injection drug users. J Clin Epidemiol 1996; 49:1149–1154. 31. Torres-Torotsa M, Rivera A, deAlarcon A, et al. Decrease in the annual frequency of infectious endocarditis among intravenous drug users in southern Spain. Enferm Infecc Microbiol 2000; 18:293–294. 32. Simberkoff MS. Narcotic-associated infective endocarditis. In: Kaplan EL, Taranta AV, eds. Infectious Endocarditis. Dallas: American Heart Association, 1977:46–58. 33. Levine DP. Infectious endocarditis in intravenous drug abusers. In: Levine DP, Sobel JD, eds. Infections in Intravenous Drug Abusers. New York: Oxford University Press, 1991:251–285. 34. Baddour LM. Twelve year review of recurrent native infective endocarditis: disease of the modern antibiotic era. Rev Infect Dis 1988; 10:1163–1170. 35. Hubbell G, Cheitlin MD, Rappaport E. Presentation, management and follow-up evaluation of infective endocarditis in drug addicts. Am Heart J 1981; 102:85–94. 36. Gouello JP, Asfar P, Brenet O, et al. Nosocomial endocarditis in the intensive care unit: an analysis of 22 cases. Crit Care Med 2000; 28:377–382. 37. Fernandez-Guerrero ML, Verdejo C, Azofra J, et al. Hospital-acquired infectious endocarditis not associated with cardiac surgery: an emerging problem. Clin Infect Dis 1995; 20:16–23. 38. Fowler VG Jr., Sanders LL, Kong LK. Infective endocarditis due to Staphylococcus aureus: 59 prospectively identify cases with follow-up. Clin Infect Dis 1999; 28:106–114. 39. Safar N, Kluger D, Maki D. A review of risk factors for catheter related bloodstream infection caused by percutaneously inserted noncuffed central venous catheters. Medicine 2002; 81:466–474. 40. Terpenning MS. Infective endocarditis. Geriatr Med 1992; 8:903–912. 41. Wilcox PA, Rayner BL, Whitelaw DA. Community-acquired Staphylococcus aureus bacteremia in patients who do not abuse intravenous drugs. Q J Med 1990; 91:41–47. 42. Garvey GJ, Neu HC. Infective endocarditis-an evolving disease. Medicine (Baltimore) 1978; 57:105–127. 43. Weinstein L, Schlesinger JJ. Pathanatomic, pathophysiologic and clinical correlations in endocarditis. N Engl J Med 1974; 29:832–837, 1122–1126. 44. Weinstein L, Brusch JL. Prophylaxis. In: Weinstein L, Brusch JL, eds. Infective Endocarditis. New York: Oxford University Press, 1995:322–337. 45. Watanakunakorn C, Burkert T. Infective endocarditis in a large community teaching hospital, 1980–1990. A review of 210 episodes. Medicine 1993; 72:90–102. 46. Weinstein L, Brusch JL. Microbiology of infective endocarditis and clinical correlates: gram-positive organisms. In: Weinstein L, Brusch JL, eds. Infective Endocarditis. New York: Oxford University Press, 1995:35–72. 47. Baddour LM, Bisno AL. Infective endocarditis complicating mitral valve prolapse: epidemiologic, clinical and microbiologic aspects. Rev Infect Dis 1986; 8:117–137. 48. Weinstein L, Brusch JL. Clinical manifestations of native valve endocarditis. In: Weinstein L, Brusch JL, eds. Infective Endocarditis. New York: Oxford University Press, 1996:165–193. 49. Gallagher P, Natanakunakora C. Group B streptococcal endocarditis: report of seven cases and review of the literature 1962–1985. Rev Infect Dis 1986; 8:175. 50. Petti CA, Fowler VG. Staphylococcus aureus bacteremia and endocarditis. In: Durack DT, ed. Infective Endocarditis. Philadelphia: WB Saunders, 2002:413–435. 51. Chambers HF. The changing epidemiology of Staphylococcus aureus. Emerg Infect Dis 2001; 7:178–182. 52. Rupp ME, Archer GL. Coagulase-negative Staphylococcus: pathogens associated with medical progress. Clin Infect Dis 1994; 19:231–245.

256

Brusch

53. Baddour LM, Phillips TN, Bisno AL. Coagulase-negative staphylococcal endocarditis: occurrence in patients with mitral valve prolapse. Arch Intern Med 1986; 146:119–121. 54. Cohen PS, Maguire JH, Weinstein L. Infective endocarditis caused by gram-negative bacteria: a review of the literature, 1945–1977. Prog Cardiovasc Dis 1980; 22:205–242. 55. Snyder N, Atterbury CE, Correia JP, et al. Increased concurrence of cirrhosis and bacterial endocarditis. Gastroenterology 1977; 33:1107–1112. 56. Ellner JJ, Rosenthal MS, Lerner PI, et al. Infective endocarditis and slow-growing fastidious gram negative bacteria. Medicine 1979; 58:145–158. 57. Baddour G, Mayer J, Henry B. Polymicrobial endocarditis in the 1980s. Rev Infect Dis 1991; 13:963–970. 58. Martin GS, Maninno DM, Eaton S, et al. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003; 348:1546–1554. 59. Leaf H, Simberkoff MS. Fungal endocarditis. In: Horstkotte D, Bodnar E, eds. Infective Endocarditis. London: ICR Publishers, 1991:118–134. 60. McLeod R, Remington JS. Fungal endocarditis. In: Rahimtoola SH, ed. Infective Endocarditis. New York: Grune and Stratton, 1978:211–290. 61. Drexler L, Rytel M, Keelan M, et al. Aspergillus terreus infective endocarditis on a porcine heterograft valve. Thorac Cardiovasc Surg 1980; 79:269–274. 62. Mayer AR, Brown A, Weintraub RA, et al. Successful medical therapy for endocarditis due to Candida parapsilosis: a clinical and epidemiologic study. Chest 1978:546–549. 63. Pesanti EL, Smith IM. Infective endocarditis with negative blood cultures: an analysis of 52 cases. Am J Med 1979; 66:43–46. 64. Weinstein L, Brusch JL. Pathoanatomical, pathophysiological and clinical correlations. In: Weinstein L, Brusch JL, eds. Infective Endocarditis. New York: Oxford University Press, 1996:138–164. 65. Moreillon P, Overholser CD, Malinverni R, et al. Predictors of endocarditis in isolates from cultures of blood following dental extractions in rats with periodontal disease. J Infect Dis 1980; 157:990–995. 66. Scheld WM, Valone JA, Sande MA. Bacterial adherence in the pathogenesis of endocarditis. J Clin Invest 1978; 61:1394–1404. 67. Yeaman MR. The role of platelets in antimicrobial host defense. Clin Infect Dis 1997; 25:951–958. 68. Aly R, Shinefield H. Role of teichoic acid in the binding of Staphylococcus aureus to normal epithelial cells. J Infect Dis 1980; 157:141–144. 69. Tompkins DC, Blackwell LJ, Hatcher VB, et al. Staphylococcus aureus proteins that bind to human endothelial cells. Infect Immun 1992; 60:965–969. 70. Hammill RJ. Role of fibronectin in infective endocarditis. Rev Infect Dis 1987; 9(suppl 4): S360–S371. 71. Drake T. Staphylococcus aureus induced tissue factor expression in cultured valve endothelium. J Infect Dis 1980; a57:749–753. 72. Sheagren J. Staphylococcus aureus, the persistent pathogen. N Engl J Med 1984; 310: 1368–1574. 73. Christensen GD, Bisno AL, Parisi B, et al. Nosocomial septicemia due to multiply antibiotic-resistant Staphylococcus epidermidis. Ann Intern Med 1982; 96:1–10. 74. Cristina AB. Biomaterial-centered infection: microbial adhesions versus tissue integration. Science 1987; 37:1588–1595. 75. Gotz F, Georg P. Colonization of medical devices by coagulase-negative staphylococci. In: Waldvogel FA, Bisno AL, eds. Infections Associated with Indwelling Medical Devices. Washington, D.C.: ASM Press, 2000:55–88. 76. Drake TA, Rodgers GM, Sande MH. Tissue factor is a major stimulus of vegetation formation in enterococcal endocarditis in rabbits. J Clin Invest 1984; 70:1750–1753. 77. Stein DS, Nelson KE. Endocarditis due to nutritionally deficient streptococci: therapeutic dilemma 1987; 9:908–916.

Infective Endocarditis and Its Mimics in the Critical Care Unit

257

78. Doring J, Maier M, Mueller E. Virulence factors of Pseudomonas aeruginosa. Antimicrob Agents Chemother 1987; 39:136–148. 79. Young F. Human immunity in Pseudomonas aeruginosa: in-vitro interaction of bacteria, polymorphonuclear leukocytes and serum factors. J Infect Dis 1972; 26:257–263. 80. Burrig KL, Schute Terhau, Sen J, et al. Special role of the endocardium in the pathogenesis of endocarditis. In: Horskotte D, Bodnar E, eds. Infective Endocarditis. London: ICR Publishers, 1991:3–9. 81. Garrison PK, Freedman LR. Experimental endocarditis. I. Staphylococcal endocarditis in rabbits resulting from placement of a polyethylene catheter in right side of the heart. J Biol Med 1978; 22:394–410. 82. Freedman LR, Valone J Jr. Experimental infective endocarditis. Prog Cardiovasc Dis 1979; 22:169–180. 83. Durack DT, Beeson PM, Petersdorf RG. Experimental bacterial endocarditis. III. Production and progress of the disease in rabbits. Br J Exp Pathol 1972; 50:50–53. 84. Rowley KM, Cluff KS, Smith GJW. Right-sided infective endocarditis as a consequence of a flow directed pulmonary artery catheterization. N Engl J Med 1984; 311: 1152–1156. 85. Greene JF, Fitzwater JE, Clemmer TP. Septic endocarditis and indwelling pulmonary artery catheters. JAMA 1975; 33:891–897. 86. Rodbard S. Blood velocity and endocarditis. Circulation 1963; 27:18–28. 87. Everett ED, Hirschmann JV. Transient bacteremia and endocarditis prophylaxis. A review. Medicine (Baltimore) 1977; 56:61–77. 88. Loesche WJ. Indigenous human flora and bacteremia. In: Kaplan EL, Taranta AV, eds. Infective Endocarditis. Dallas: American Heart Association, 1977:40–45. 89. Baskin RW, Rosenthal A, Bruitt BA. Acute bacterial endocarditis, a silent source of sepsis in the burn patient. Ann Surg 1976; 184:618–625. 90. Raad H, Bodey GP. Infectious complications of indwelling vascular catheters. Clin Infect Dis 1992; 15:197–210. 91. Durack DT. Prevention of infective endocarditis. N Engl J Med 1995; 332:38–44. 92. Sheretz RJ. Pathogenesis of vascular catheter infection. In: Infections Associated with Indwelling Medical Devices. Washington, D.C.: ASM Press, 2000:111–125. 93. Sitges-Serra A, Linares J, Garau J. Catheter sepsis: the clue is the hub. Surgery 1985; 97:355–357. 94. Sheth NK, Franson TR, Rose HD. Colonization of bacteria on polyvinylchloride Teflon intravascular catheter in hospitalized patients. J Clin Microbiol 1983; 18:1061–1063. 95. Raad I, Costerton JW, Sabharwar U, et al. Central venous catheters (CVC) studied by quantitative cultures and scanning electron microscopy (SEM): the importance of luminal colonization. Program and Abstracts of the 31st Interscience Conference on Antimicrobial Agents and Chemotherapy. Abstract 450 1991. Washington, D.C.: American Society of Microbiology, 1991. 96. Jarvis WR. Nosocomial outbreaks: The Center for Disease Control’s hospital infections program experience. 1980–1991. Am J Med 1991; 91(suppl B):101S–106S. 97. Finland M, Barnes MW. Changing etiology of bacterial endocarditis in the antibacterial era: experiences in The Boston City Hospital 1933–1955. Ann Intern Med 1970; 72: 341–348. 98. Snydman DR, Reidy MD, Perry LK, et al. Safety of changing intravenous (IV) administration sets containing burettes at longer than 48 hour intervals. Infect Control 1987; 8:113–116. 99. Maki DG, Rhame FS, Mackel DC, et al. Nationwide epidemic of septicemia caused by contaminated intravenous products. Am J Med 1976; 60:471–485. 100. Johnson JD, Raff MJ, Barnwell PA. Splenic abscess complicating infectious endocarditis. Arch Int Med 1983; 143:906–912. 101. Wilson WR, Lie JT, Houser OW, et al. The management of patients with mycotic aneurysm. Curr Clin Topics Infect Dis 1981; 2:151–181.

258

Brusch

102. Lerner PI, Weinstein L. Infective endocarditis in the antibiotic era. N Engl J Med 1966; 274:199–206, 259–266, 323–331, 388–393. 103. Weinstein L. Life-threatening complications of infective endocarditis and their management. Arch Int Med 1988; 146:953–957. 104. Roberts WC, Buchbinder NA. Healed left-sided infective endocarditis: a clinicopathological study of 59 patients. Am J Cardiol 1976; 40:876–883. 105. Ghosh PK, Miller HJ, Vidne BA. Mitral obstruction in bacterial endocarditis. Br Heart J 1985; 53:340–344. 106. Crawford HH, Badke FR, Amon KW. Effectively undisturbed pericardium on left ventricular size and performance during acute volume loading. Am Heart J 1983; 105: 267–270. 107. Fang G, Keys T, Gentry L. Prosthetic valve endocarditis resulting from nosocomial bacteremia. Ann Intern Med 1993; 119:560–567. 108. Horskotte D. Prosthetic valve endocarditis. In: Horskotte D, Bodnar E, eds. Infective Endocarditis. London: ICR Publishers, 1991:233–261. 109. Arber, Pras E, Coopperman Y, et al. Pacemaker endocarditis. Report of 44 cases and review of the literature. Medicine (Baltimore) 1994; 73:299–315. 110. Cherubin CE, Sapira JD. The medical complications of drug addiction and medical assessment of intravenous drug users: 25 years later. Ann Intern Med 1993; 119: 1017–1028. 111. Werner BS, Schulz R, Fuchs JB, et al. Infective endocarditis in the elderly in the era of transesophageal echocardiography: clinical features and prognosis compared with younger patients. Am J Med 1976; 100:90–97. 112. Weinstein L, Rubin R. Infective endocarditis-1973. Prog Cardiovasc Dis 1973; 16:239–274. 113. Libman E, Friedberg CK. Subacute Bacterial Endocarditis. Oxford: Oxford University Press, 1948. 114. Weinstein L. Infective endocarditis: past, present and future. J R Coll Phys Lond 1972; 6:161–163. 115. Churchill M, Geraci J, Hunder G. Musculoskeletal manifestations of bacterial endocarditis. Ann Intern Med 1977; 87:755–762. 116. Weinstein MP, Towns ML, Quartey SM, et al. The clinical significance of positive blood cultures in the 1990s: a prospective comprehensive evaluation of the microbiology, epidemiology, and outcome of bacteremia and fungemia in adults. Clin Infect Dis 1997; 24:584–602. 117. Weinstein L. Infective endocarditis. In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine. 3rd ed. Philadelphia: WB Saunders, 1988:113. 118. Bates D, Lee TH. Rapid classification of positive blood cultures; prospective validation of a multivariate algorithm. JAMA 1992; 267:1962–1966. 119. Weinstein MP. Current blood culture methods in systems: clinical concepts, technology and interpretation of results. Clin Infect Disease 1996; 23:40–46. 120. Miller M, Casey J. Infective endocarditis: new diagnostic techniques. Am Heart J 1978; 96:123–130. 121. Safar N, Fine JP, Maki D. Meta-analysis: evidence for diagnosing intravascular devicerelated bloodstream infection. Ann Intern Med 2005; 142:451–466. 122. Murray PR, Traynor P, Hopson D. Critical assessment of blood culture techniques: analysis of recovery of complicated facultative anaerobes, strict anaerobic bacteria and fungi in aerobic and anaerobic blood culture bottles. J Clin Microbiol 1992; 30:1462–1468. 123. Von Scoy RE. Culture-negative endocarditis. Mayo Clin Proc 1982; 57:149–154. 124. Keefer CS. Subacute bacterial endocarditis: active cases without bacteremia. Ann Intern Med 1937; 11:714–734. 125. Barbari EF, Cockerill FR, Steckelburg JM. Infective endocarditis due to unusual or fastidious microorganisms. Mayo Clin Proc 1997; 72(6):532–542. 126. Vansdoy RE. Culture negative endocarditis. Mayo Clin Proc 1982; 57:149–156.

Infective Endocarditis and Its Mimics in the Critical Care Unit

259

127. Lichtlen P. General principles of conservative treatment of infective endocarditis. In: Horstkotte D, Bodnar E, eds. Infective Endocarditis. London: ICR Publishers, 1991: 85–92. 128. Bates D, Goldmann L, Lee TH. Contaminant blood cultures and resource utilization. JAMA 1991; 265:365–369. 129. Hoen B, Selton-Suty C, Lacassin F, et al. Infective endocarditis in patients with negative blood culture: analysis of 8 cases in a one year nationwide survey in France. Clin Infect Dis 1995; 20:501–506. 130. Arnette N, Roberts SI. Valve ring abscesses in active infective endocarditis. Circulation 1976; 54:140–145. 131. Oates E, Sarno RC. Detection of bacterial endocarditis with indium-III labeled leukocytes. Clin Nucl Med 1988; 13:691–693. 132. Lindner JR, Case RA, Dent JM, et al. Diagnostic value of echocardiography in suspected endocarditis. An evaluation based upon the pretest probability of disease. Circulation 1996; 93:730–740. 133. Lowry RW, Zogbhi WA, Baker WB, et al. Clinical impact of transesophageal echocardiography in the diagnosis and management of infective endocarditis. Am J Cardiol 1994; 73:1089–1091. 134. Chirillo F, Bruni A, Giujusa T, et al. Echocardiography in infective endocarditis: reassessment of the diagnostic criteria of vegetation as evaluated from the precordial and transesophageal approach. Am J Card Imaging 1995; 9:174–179. 135. Roe MT, Abramson MA, Li J, et al. Clinical information determines the impact of transesophageal echocardiography on the diagnosis of infective endocarditis by the Duke criteria. Am Heart J 2000; 139:945–953. 136. Cheitlin MD, Armstrong WF, Aurigemma GP, et al. ACC/AHA/ASE 2003 guideline update for the clinical application of echocardiography: summary article. A report of the American College of Cardiology/American Heart Association task force on practice guidelines. Circulation 2003; 108:1146–1153. 137. Ascher DP, Shoupe BA, Maybee D, et al. Persistent catheter-related bacteremia: clearance with antibiotics and urokinase. J Pediatr Surg 1993; 28:628–635. 138. Durack DT, Beeson PB. Experimental bacterial endocarditis. Part II. Survival of bacteria in the endocarditis vegetation. Br J Exp Pathol 1972; 53:50–53. 139. Sanfillipo AJ, Picard MH, Newell JB, et al. Echocardiographic assessment of patients with infectious endocarditis: prediction of risk for complications. J Am Coll Cardiol 1991; 18:1191–1199. 140. Rohmann S, Erbel R, Gorge T, et al. Clinical relevance of vegetation localization by transesophageal echocardiography in infective endocarditis. Eur Heart J 1992; 12:446–452. 141. Omari B, Shapiro S, Gintzon L, et al. Predictive risk factors for perivalvular extension of native valve endocarditis. Clinical and echocardiographic analyses. Chest 1989; 96(6): 1273–1279. 142. Jeang M, Fuenfes F, Gately A, et al. Aortic root abscess: initial experience using magnetic resonance imaging. Chest 1986; 89(4):613–615. 143. Goldenberger D, Kunzli A, Vogt P, et al. Molecular diagnosis of bacterial endocarditis by broad-range PCR amplification and direct sequencing. J Clin Microbiol 1997; 35:2733–2739. 144. Li JS, Sexton DJ, Mick N, et al. Proposed modifications to the Duke criteria for diagnosis of infective endocarditis. Clinic Infect Dis 2000; 30:633–644. 145. Rosenblum G, Carsons S. Mimics of endocarditis. In: Cunha B, ed. Infectious Diseases in Critical Care Medicine. New York: Marcel Dekker, 1998:435–434. 146. Maksimowicz-McKinnon K, Mandell BF. Understanding valvular heart disease in patients with systemic autoimmune diseases. Cleveland Clinic J Med 2004; 11:881–885. 147. Brusch JL. Cardiac infections in the immunosuppressed patient. In: Cunha B, ed. Infections in the Compromised Host Infectious Disease Clinics of North America. Philadelphia: WB Saunders Co, 2001:613–638. 148. Fisher J. Cardiac myxoma. Cardiovasc Rev Rep 1983; 9:1195–2001.

260

Brusch

149. Bauernschmitt R, Jakob HG, Vahl C-F, et al. Operation for infective endocarditis: results after implementation of mechanical valves. Ann Thorac Surg 1998; 65:359–364. 150. Truninger K, Attenhofer Jost CH, Seifert B, et al. Long-term follow-up of prosthetic valve endocarditis: What characteristics identify patients who were treated successfully with antibiotics alone? Heart 1999; 82:714–720. 151. Bayer AS, Bolger AF, Taubert KA, et al. Diagnosis and management of infective endocarditis and its complications. Circulation 1998; 25:2936–2948. 152. Olaison L, Hogevik H, Myken P, et al. Early surgery in infective endocarditis. Q J Med 1996; 89:267–278. 153. Magilligan D. Cardiac surgery in infective endocarditis. In: Horstkotte D, Bodnar E, eds. Infective Endocarditis. London: ICR Publishers, 1991:210. 154. Wilkoff BL, Byrd CL, Love CJ, et al. Pacemaker lead extraction with the laser sheath: results of the patient lead extraction with the excimer sheath (Plexes) trial. J Am Coll Cardiol 1999; 33:1671–1685. 155. Lee BK, Crossley K, Gerding DN. The association between Staphylococcus aureus bacteremia and bacteruria. Am J Med 1978; 65:303–310. 156. Torlakovic E, Hibbs JR, Miller JS, et al. Intracellular bacteria in blood smears in patients with central venous catheters. Arch Int Med 1995; 155:1547–1553. 157. Fowler V, Li J, Core GR, et al. Role of echocardiography in evaluation of patients with Staphylococcus aureus bacteremia in 107 patients. J Am Coll 1997; 30:107–218. 158. Chang FY. A prospective multicenter study of Staphylococcus aureus bacteremia: incidence of endocarditis, risk factors for mortality and clinical impact of methicillinresistant. Medicine (Baltimore) 2003; 82:322–332. 159. Press OW, Ramsey PV, Larson EB, et al. Hickman catheter infections in patients with malignancies medicine 1984; 63:189–200. 160. Joly V, Pangon B, Vallois JM, et al. Value of antibiotic levels in serum and cardiac vegetations for predicting antibacterial effect of ceftriaxone in experimental E. coli endocarditis. Antimicrob Agents Chemother 1987; 31:1632–1635. 161. Gengo F, Schentag J. Rate of methicillin penetration into normal heart valves and experimental endocarditis lesions. Antimicrob Agents Chemother 1982; 21:456–459. 162. Fraimow HJ, Abrutyn E. Pathogens resistant to antimicrobial agents: epidemiology, molecular mechanisms and clinical management. Infect Dis Clin North Am 1995:497–530. 163. Douglas A, Moore-Gillon J, Eykyn J. Fever during treatment of infective endocarditis. Lancet 1986; 1:1341–1343. 164. Blumberg E, Robbins SN, Adimora A, et al. Persistent fever in association with infective endocarditis. Clin Infect Dis 1992; 15:980–996. 165. Sexton DJ, Spelman D. Current best practices and guidelines: assessment and management of complications in infective endocarditis. In: Durack DT, ed. Infectious Disease Clinics of North America. Philadelphia: WB Saunders Company, 2002:16507–16521. 166. Wilson W, Giuliani E, Danielson G, et al. Management of complications of infective endocarditis. Mayo Clin Proc 1982; 57:162–169. 167. Weinstein L, Brusch JL. In: Weinstein L, Brusch JL, eds. Medical Management. New York: Oxford University Press, 1996:256–304. 168. Welton DE, Young JB, Gentry WO, et al. Recurrent infective endocarditis: analysis of predisposing factors and clinical features. Am J Med 1979; 66:932–939. 169. Francioli P, Ruch W, Stamboulian D. Treatment of streptococcal endocarditis with a single daily dose of ceftriaxone and netilmicin for 14 days: a prospective multicenter study. Clin Infect Dis 1995; 21:1406–1410. 170. Doern GV, Ferraro MJ, Bruggermann AB, et al. Emergence of high rates of antimicrobial resistance among viridans streptococci in the United States. Antimicrob Agents Chemother 1995; 39:2243–2247. 171. Pulliam L, Inokuchi S, Hadley WK, et al. Penicillin tolerance in experimental streptococcal endocarditis. Lancet 1979; 2:957–961.

Infective Endocarditis and Its Mimics in the Critical Care Unit

261

172. Eliopoulos GM. Antibiotic resistance in enterococcus species: an update. In: Remington J, Schwarz M, eds. Current Clinical Topics in Infectious Diseases. Cambridge, MA: Blackwell Scientific, 1996:21–51. 173. Eliopoulos GM. Aminoglycoside resistant enterococcal endocarditis. Med Clin North Am 1993; 17:117–172. 174. Fantin B, Carbon C. Importance of the aminoglycoside dosing regimen in the penicillinnetilmicin combination for treatment of Enterococcus faecalis-induced experimental endocarditis. Antimicrob Agents Chemother 1990; 34:2387–2391. 175. Wilson WR, Karchmer AW, Dsajani AS, et al. Antibiotic treatment of adults with infective endocarditis due to streptococci, enterococci, staphylococci and HACEK microorganisms. JAMA 1995; 274:1705–1714. 176. Thompson RL, Lavin B, Talbot GH. Endocarditis due to vancomycin-resistant Enterococcus faecium in an immunocompromised patients: cure by administering combination therapy with Quinpristin/dalfopristin and high dose ampicillin. South Med J 2003; 96:818–820. 177. Eliopoulos GM, Thavin-Eliopoulos C, Moellering RC Jr. Contribution of animal models in the search for effective therapy for endocarditis due to enterococci with high-level resistance to gentamicin. Clin Infect Dis 1992; 15:58–62. 178. Karchmer AW. Staphylococcus aureus and vancomycin: the sequel. Ann Intern Med 1991; 115:739–741. 179. Nannini EC, et al. Relapse of type A beta-lactamase-producing Staphylococcus aureus native valve endocarditis cefazolin therapy: revisiting the issue. Clin Infect Dis 2003; 37:1194–1198. 180. Korzeniowski O, Sande MH. The national collaborative endocarditis study group combination antimicrobial therapy for Staphylococcus aureus endocarditis in patients addicted to parenteral drugs and in non-addicts: a prospective study. Ann Intern Med 1982; 92:619–624. 181. Chambers HF. Short-course combination and oral therapies of Staphylococcus aureus endocarditis. Med Clin North Am 1993; 7:69–80. 182. Karchmer AW. Infections of prosthetic heart valves. In: Waldvogel F, Bisno Al, eds. Infections Associated with Indwelling Medical Devices. Washington, D.C.: American Society for Microbiology, 2000:145–172. 183. Chuard C, Herrmann M, Roehner P, et al. Treatment of experimental foreign body infections caused by methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemother 1990; 34:2312–2317. 184. Stevens DL, Herr D, Lampiris H, et al. Linezolid versus vancomycin for the treatment of methicillin-resistant Staphylococcus aureus infections. Clin Infect Dis 2002; 34:1481–1490. 185. Howden BP, Ward PB, Charles PG, et al. Treatment outcomes for serious infections caused by methicillin-resistant Staphylococcus aureus with reduced vancomycin susceptibility. Clin Infect Dis 2004; 38:521–528. 186. Jacquiline C, Batard E, Perez L, et al. In vivo efficacy of continuous infusion versus intermittent dosing of linezolid compared to vancomycin in a methicillin-resistant Staphylococcus aureus rabbit endocarditis model. Antimicrob Agents Chemother 2002; 46:3706–3711. 187. Sperber SJ, Levine JF, Gross PA. Persistent MRSA bacteremia in a patient with low Linezolid levels. Clin Infect Dis 2003; 36:675–676. 188. Sakoulas G, Eliopoulos GM, Alder J, et al. Efficacy of daptomycin in experimental endocarditis due to methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemother 2003; 47:1714–1718. 189. Ellis ME, Al Abdely H, Sandridge A, et al. Fungal endocarditis: evidence in the world literature 1965–1995. Clin Infect Dis 2001; 32:50–62. 190. Walsh T, Pappas P, Winston DJ, et al. Voriconazole versus liposomal amphotericin B for empirical antifungal therapy. N Engl J Med 2002; 366:1745–1747.

262

Brusch

191. Mora-Duarte J, Betts R, Rotstein C, et al. Comparison of caspofungin amphotericin B for invasive candidiasis. N Engl J Med 2002; 347:2020–2029. 192. Anaisse EJ, Bishara AB, Solomkin JS. Fungal infection. In: Souba WW, Fink MP, Jurkovich GJ, et al eds. New York: Web Professional Publishing 2005:1486–1487. 193. Dajani AS, Taubert KA, Wilson W, et al. Prevention of bacterial endocarditis: recommendations by the American Heart Association. Circulation 1997; 96:358–366. 194. Centers for Disease Control and Prevention. Guidelines for the prevention of intravascular catheter-related infections. Morb Mortal Wkly Report 2002; 51:1–29.

12 Acute Myocarditis and Acute Pericarditis in the Critical Care Unit Jason M. Lazar Department of Cardiology, State University of New York Downstate Medical Center, Brooklyn, New York, U.S.A.

Diane H. Johnson and Burke A. Cunha Infectious Disease Division, Winthrop-University Hospital, Mineola, and State University of New York School of Medicine, Stony Brook, New York, U.S.A.

ACUTE MYOCARDITIS Introduction Acute myocarditis is a common sequelae of systemic infection by many organisms. It is estimated to occur in 5% to 15% of common infections. Acute myocarditis develops from direct infection or after infection with the heart as a target of immune injury (1). Most cases are subclinical, but acute myocarditis can progress to congestive heart failure and death. It may also preset with ventricular arrhythmias or resemble acute myocardial infarction (MI). Myocarditis can become a chronic progressive disease and is estimated to account for 10% to 20% of cases of dilated cardiomyopathy (2). This chapter provides an overview of the clinical course of acute infectious myocarditis. Infectious Causes The clinical syndrome of myocarditis was first described in the mid-1800s in patients with mumps and epidemic pleurodynia. Myocarditis was encountered during the influenza A pandemic during the first part of the century. Although the causative agent of myocarditis is not always identified, most cases of infectious myocarditis in the United States and Europe are viral in origin (3–5). The group B coxsackieviruses are responsible for most cases of documented human disease. Other enteroviruses, including the coxsackie A and echoviruses, are important causes of myocarditis as well (4). Historically, poliovirus was a cause of myocarditis; however, its incidence has markedly declined due to widespread vaccination efforts. These enteroviruses are RNA viruses that attach to receptors on the cardiac myocyte and cause cell destruction and the subsequent initiation of a host immune response. Other commonly encountered viruses that may result in myocarditis clinically include: varicella, cytomegalovirus (CMV), Epstein-Barr virus (EBV), 263

264

Lazar et al.

rubeola, rubella, hepatitis B, and adenovirus (6). Arboviruses, such as dengue and the rabies virus, are also causes of myocarditis (5,7). Currently, HIV infection is often associated with cardiomyopathy. Bacterial pathogens are responsible for some cases of myocarditis and reach the myocardium by direct hematogenous spread of microorganisms or from contiguous spread from an infected heart valve. Gonococci, Meningococci, Brucella, Salmonella, Staphylococci, and Streptococci have all been reported to cause bacterial myocarditis (5). Bacteria such as Corynebacterium diphtheriae and Clostridium perfringens elaborate toxins that can damage the myocardial tissue (8,9). Some organisms causing atypical pneumonias, such as Mycoplasma pneumoniae, Legionella pneumophila, Chlamydia pneumoniae, and Chlamydia psittaci, are unusual but are known causes of myocarditis. Lyme disease, which is caused by the spirochete Borrelia burgdorferi, is known to cause myocarditis, which usually manifests as conduction disturbances. Infections with rickettsia commonly cause myocarditis as well, e.g., Rocky Mountain spotted fever and scrub typhus (10). In South America, myocarditis secondary to Chagas disease caused by Trypanosoma cruzzi is fairly common. The other trypanosome species, T. gambiense and T. rhodesiense, that cause African sleeping sickness can infect the heart as well. Trichinella spiralis, Toxoplasma gondii, and Echinococcus are causes of myocarditis in developing nations (11). Disseminated fungal infections such as aspergillosis, cryptococcosis, and candidiasis have all been reported to result in myocarditis, with the majority occurring in immunocompromised individuals (12). The infectious causes of myocarditis and the associated clinical and laboratory features are listed in Table 1. Clinical Presentation The majority of cases of acute infectious myocarditis are presumed to be asymptomatic, leaving its incidence and course ill defined. Alternatively, acute myocarditis may result in congestive heart failure, chest pain mimicking acute MI, and/or arrhythmias. Because these conditions are more commonly associated with primary myocardial disease, the diagnosis of acute myocarditis poses a challenge. Congestive Heart Failure The overall incidence of myocarditis in acute heart failure is unknown. It has been reported to accompany between 4% and 80% of cases of acute and chronic cardiomyopathy collectively (13), but occurs more frequently in patients with symptoms of shorter duration. In patients presenting with signs and symptoms of congestive heart failure, active myocarditis should be considered when there is a young patient age, absence of cardiac history, and the onset of symptoms during or immediately following a systemic or viral illness. The illness may present with flu-like symptoms of respiratory or gastrointestinal nature including fever, chills, cough, coryza, myalgia, pharyngitis, anorexia, and diarrhea. Of note, cardiac involvement is more likely to occur in patients reporting myalgias during the prodromal illness (14). The signs and symptoms of heart failure due to myocarditis are similar to other causes. Dyspnea, orthopnea, paroxysmal nocturnal dyspnea, and fatigue are manifestations of left ventricular (LV) dysfunction. Abdominal discomfort related to hepatomegaly and peripheral edema result from right ventricular failure. The duration of symptoms is brief (less than three months) in over two-thirds of patients with biopsy-proven myocarditis (13). In addition, substernal or precordial chest pain (Text continued on page 274)

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

265

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests Causes Myocarditis Viral Group B

Associated clinical features Fever

Coxsackieviruses

Upper respiratory tract symptoms Chest pain (pleurodynia) Skin rash

Echoviruses

Fever Upper respiratory tract symptoms Loose stool or diarrhea Skin rash

Influenza A

Fever

Influenza B

Headache Myalgias Eye pain Nonproductive cough Sore throat

Adenovirus

Rhinorrhea Pharyngitis Tracheitis Pneumonia Conjunctivitis Cervical adenitis Fever Cough Coryza Conjunctivitis Koplik spots Descending blotchy maculopapular rash that becomes confluent Fever Mild conjunctivitis Posterior cervical/occipital adenopathy Arthralgia Pink, macular/papular rash beginning on face spreading downward Palatal petechias (Forscheimer spots)

Measles (rubella)

Rubella

Diagnostic tests Isolation of virus from throat or stool " Titer IgM coxsackie B antibody or 4 rise in IgG antibody titer (CF) between acute and convalescent sera Isolation of virus or viral proteins from myocardial biopsy " Titer IgG echovirus antibody or four fold rise in IgG antibody titer (CF) between acute and convalescent sera Isolation of influenza virus from nasopharynx, throat or sputum Isolation of influenza virus from nasopharynx, throat, or sputum Four-fold rise in IgG antibody titer between acute and convalescent sera (EIA, IFA, CF, HIA) Isolation of virus or viral proteins from myocardium Culture of adenovirus from pharynx sputum, conjunctiva, urine Four-fold rise in IgG adenovirus antibody titer between acute and convalescent sera (EIA, IFA) In early phase, multinucleated giant cells seen in stained nasal secretions, sputum Leukopenia Measles-specific IgM antibody (HIA, EIA, IFA, CF) or Four-fold rise in IgG titer between acute and convalescent sera Isolation of virus from nasopharynx Rubella specific IgM antibody (EIA or HAI or four fold rise in titer between rubella-specific IgG antibodies (EIA, HAI)

(Continued )

266

Lazar et al.

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes Varicella

Mumps

Polio

Epstein-Barr virus

Myocarditis Viral Cytomegalovirus

Dengue

Associated clinical features

Diagnostic tests

Prodromal flu-like symptoms Tzanck prep of lesionsRash initially macular, becomes, multinucleated giant cells vesicular, then pustular Basophilia Rash erupts in crops Demonstration of virus by IFA from material from a lesion Varicella specific IgG antibodies between acute and Convalescent sera (EIA, CF) " Amylase Parotitis mostly bilateral, but can be unilateral (25%) Isolation of virus from blood, Fever nasopharynx, secretions from Epididymo-orchitis in men Stensen’s duct, cerebrospinal (-30%) fluid (CSF), urine " Mumps specific IgM, or fourfold rise in IgG, IFA mumps antibody between acute and convalescent sera Fever Mild CSF pleocytosis Aseptic meningitis (predominantly lymphs), slightly Muscle pain, cramps, twitching elevated protein Asymmetric paralysis Isolation of virus from throat, stool Sensation remains intact Four-fold rise in CF antibody, EIA, or IFA titer or polio antibody Fever Mildly " LFTs Fatigue Lymphocytosis with atypical Periorbital edema lymphocytes Pharyngitis (may be exudative) Positive heterophile antibody Posterior cervical or generalized monospot test lymphadenopathy Positive IgM viral capsid antigen high titers present serum Splenomegaly 1–6 wks after onset of illness " ESR Fever Malaise, fatigue Myalgia, headache Splenomegaly Pharyngitis and cervical adenopathy are uncommon Prodromal symptoms: anorexia, nausea vomiting, fatigue arthralgia, myalgia, pharyngitis, cough, possible rash Distortion of taste Icterus Tender hepatomegaly Saddle-back fever curve Mild conjunctivitis Severe headache

Mildly " LFTs Lymphocytosis with atypical lymphocytes Latex agglutination test for CMV antibody Four fold rise in CMV IgG antibody titer PCR evaluation of blood Markedly elevated LFSs Initial lymphopenia followed by lymphocytosis with atypical lymphocytes Leukopenia Virus may be isolated from blood (Continued )

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

267

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes

Lymphocytic choriomeningitis

Associated clinical features Myalgias, especially back, lower extremities Pain with eye movement Pinpoint vesicles on posterior soft palate Skin rash initially erythema, becoming morbiliform on thorax, inner arms, followed by appearance of pruritic maculopapular desquamating rash Initial nonspecific illnessoccasional lymphadenopathy or maculopapular rash resolves in 2–4 days Illness recurs with severe headache, meningitis

Rabies

Prodromal-flu–like illness Hyperactivity, hallucinations, bizarre Hypertension Salivation Paralysis

Yellow fever

Early Fever, relative bradycardia, myalgia Severe headache Lumbosacral pain Conjunctival ejection Coated tongue with red edges Later Jaundice, delirium, acidosis, shock Urethritis-usually purulent discharge, dysuria Mucopurulent cervicitis in females Arthritis Pustular skin rash on distal extremities

Bacterial Neisseria gonorrhoeae

Diagnostic tests Presence of dengue specific IgM antibody (EIA, CF, IFA) Four-fold rise in titer between acute and convalescent IgG antibody (EIA, CF, IFA)

Leukopenia, thrombocytopenia CSF: lymphocyte predominance, usually elevated CSF pressure Isolates of virus in blood (early), and CSF (late) "LCM IgM titer or four-fold rise LCM IgG titer titers (CF, IFA) between acute and convalescent sera No test available for diagnosis prior to onset of clinical disease Microscopic examination of brain tissue for Negri bodies from rabid animals Positive CF or ELISA antibody from CSF (not detected until after onset of clinical illness) Serum is positive by mouse intracerebral inoculation in less than 5 days Postmortem pathological examination of liver

Intracellular gram-negative diplococci on Gram stain of urethral, cervical secretions, synovial fluid Blood cultures in disseminated disease (Continued )

268

Lazar et al.

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes Neisseria meningitidis

Brucella

Salmonella

Myocarditis Bacterial Staphylococcus aureus Streptococcus pyogenes

Associated clinical features Fever Macular or petechial skin rashmost commonly found on axillae, flanks, wrists, ankles Meningitis Fever Arthralgia, myalgia Anorexia, weight loss Splenomegaly on 20% Fever with relative bradycardia Constipation or diarrhea Abdominal tenderness Mild hepatosplenomegaly Gastroenteritis

Elevated PMNs in CSF Cultures of blood, CSF, nasopharynx Aspirate from skin lesion

Fever

Blood cultures Organism cultures from infected site Elevated teichoic acid antibody Blood cultures

Fever Sandpaper-like rash Erythema marginatum

Corynebacterium diphtheriae

Legionella pneumophila

Arthritis Pharyngitis Subcutaneous modules Sore throat Tonsillar membrane-dirty gray, may be green or necrotic-may extend over soft palate Hoarseness, dyspnea, stridor Neuropathy with severe disease Pneumonia Relative bradycardia Abdominal pain, diarrhea Mental confusion

Mycoplasma pneumoniae

Diagnostic tests

Cough, usually nonproductive, may be accompanied by wheezing Headache Pharyngitis Minority have myringitis Erythema multiforme Other skin rashes

Blood cultures Acute and convalescent serological titers Blood cultures Isolation of organism from stool Acute and convalescent serological titers

Throat culture Elevated ASO, anti-DNAase B, antihyaluronidase Elevated ESR

‘‘Chinese letter’’ arrangement on Gram stain of membrane Culture of membrane (if diphtheria is suspected, alert lab to use selective media)

Fluorescent antibody staining of sputum Isolation of organism of sputum Four-fold rise in titer between acute and convalescent sera Elevated serum Legionella antibody titer Legionella urinary antigen Elevated cold agglutinin titer Mycoplasma IgM, IgG serology Detection of M. pneumoniae antigen in sputum Sputum cultures require special media (Continued )

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

269

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes Chlamydia pneumoniae Chlamydia psittaci

Borellia burgdorferi

Rickettsia rickettsii

Rickettsia tisutsugamushi

Mycarditis Parasitic Trypanosoma orazii

Trichinella spirulis

Associated clinical features Fever Pharyngitis with hoarseness Sinusitis History of bird exposure Fever Pharyngitis with hoarseness Splenomegaly Hepatomegaly Relative bradycardia Epistaxis History of tick exposure Bull’s-eye rash Headache Meningismus Arthralgias Often history of tick exposure Fever Headache Myalgia Nausea, vomiting, abdominal pain Rash-initially maculopapular around wrists, ankles, becomes petechial Papule, which ulcerates to form an eschar at site of chigger bite High temperature Severe headache Myalgias Tender lymphadenopathy in region of bite Conjunctival injection Ocular pain

Chagoma—indurated, crythematous area on skin if parasite enters there Romana sign—periorbital/ palpebral edema Fever Malaise Hepatosplenomegaly Generalized lymphadenopathy Most infections asymptomatic Diarrhea, nausea, vomiting, abdominal pain

Diagnostic tests Culture of pharyngeal swab C. pneumoniae IgM, IgG serology Four-fold rise in serological titer

Serological testing including Western blot

Isolation of organism from blood Demonstration of organism from biopsy of skin lesion Serological testing

Weil-Felix–antibodies to OX-K in 50% Four-fold rise in 50% serological titer

Detection of parasites in buffy coat/blood Detection of parasites in lymph nodes, bone marrow aspirate, pericardial fluid Serological testing for T. cructi IgG, IgM

Eosinophilia Low ESR (approx. 0) (Continued )

270

Lazar et al.

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes

Associated clinical features Periorbital edema, subconjunctival hemorrhage Fever

Toxoplasma gondii

Fungal (Note: most common in immunocompromized biosis as a result of disseminated infection) Cryptococcus neoformans

Aspergillus species

Candida species

Pericarditis Viral Group B Coxsackieviruses

Myositis Asymptomatic cervical lymphadenopathy Occasionally mononucleosis-like illness

Diagnostic tests Four-fold rise between acute and convalescent serology Trichinella cysts seen on muscle biopsy T. gondii IgM/IgG serology Toxoplasma cysts in tissue from biopsy specimen

Culture of organism from CSF, Headache, dizziness, sputum, urine, blood somnolence, impaired memory, judgement with CNS disease Dull chest pain, cough, dyspnea India ink preparation of CSF may with pulmonary disease assist in presemptive diagnosis Skin lesions—popular, pastular, Visualization of organism in or nodular histological specimen Detection of cryptococcal polysaccharide antigen from serum and/or CSF High fever Isolation of organism from blood, Consolidation on chest X-ray CSF, bone marrow (rarely positive) Rarely cerebral infarct due to Isolation from sputum in CNS invasion appropriate clinical sening Necrotizing skin lesions Visualization of organism in histological specimen Fever Isolation of organism from blood Fluffy infiltrates on retina Visualization of organism in eye examination histological specimen Nodular skin lesions

Fever Upper respiratory tract symptoms Chest pain (pleurodynia) Skin rash

Isolation of virus from throat, stool " Coxsackie B IgM titer or four-fold rise in antibody titer between acute and convalescent sera (Continued )

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

271

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes Echoviruses

Influenza A

Associated clinical features Fever Upper respiratory tract symptoms

Diarrhea Skin rash Fever

Influenza B

Headache Myalgia Eye pain Nonproductive cough Sore throat

Adenovirus

Rhinorrhea Pharyngitis Tracheitis Pneumonia Conjunctivitis Cervical adenitis Parotitis, mostly bilateral, but can be unilateral (25%) Fever Epididymo-orchitis in men (30%)

Mumps

Pericarditis Viral Epstein-Barr virus

Cytomegalovirus

Fever Periorbital edema Pharyngitis (may be exudative) Posterior cervical or generalized Lymphadenopathy Splenomegaly

Fever Melaise, fatigue Myalgia, headache Splenomegaly Pharyngitis and cervical

Diagnostic tests Isolation of virus or viral proteins from pericardium " Echovirus IgM titer or fourfold rise in IgG antibody titer (CF) between acute and convalescent sera Isolation of virus or viral proteins from pericardium Isolation of influenza virus from nasopharynx, throat, or sputum Four-fold rise in IgG antibody titer between acute and convalescent sera (EIA, IFA, CF, HIA) Isolation of virus or viral proteins from pericardium Culture of adenovirus from pharynx sputum, conjunctive, urine Four-fold rise in adenovirus IgG antibody titer between acute and convalescent sera " amylase Isolation of virus from blood, nasopharynx, secretions from Stensen’s duct, CSF, urine " mumps-specific IgM or fourfold rise in IgG, IFA, mumps antibody between acute and convalescent sera

Mildly " LFTs Lymphocytosis with atypical lymphocytes Positive monospot antibody test Positive IgM viral capsid antigen high titers present in serum 1–6 wks after onset of illness, " ESR Mildly " LFTs Lymphocytosis with atypical lymphocytes Latex agglutination test for CMV antibody (Continued )

272

Lazar et al.

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes

Associated clinical features

Diagnostic tests

adenopathy are uncommon

Hepatitis B

Varicella zoster virus

Bacterial Mycoplosma pneumoniae

Legionella pneumophila

Chlamydia pneumoniae Borella burgdorferi

Actinomyces spp.

Four-fold rise in IgG antibody titer, PCR of blood Prodromal symptoms: anorexia Markedly elevated LFTs nausea, vomiting, fatigue, Initial lymphopenia followed by lymphocytosis with atypical arthralgia, myalgia, pharyngitis, cough, possible rash lymphocytes Distortion of taste Detection of HBsAg in serum Iaterus Tender hepatomegaly Prodromal flu-like symptoms Tzanck prep of lesions shows Rash initially macular, becomes multinucleated giant cells vesicular, then pustular Basophilia Rash erupts in crops Demonstration of virus by DFA from material from a lesion Varicella specific IgM antibody (EIA) Four-fold rise in varicella-specific IgG antibodies between acute and convalescent sera (EIA, CF) Cough, usually nonproductive, may be accompanied by wheezing Headache Pharyngitis Minority have myringitis Erythema multiforme Other skin rashes Pneumonia Relative bradycardia Abdominal pain, diarrhea Mental confusion

Fever Pharyngitis with hoarseness Sinusitis History of tick exposure Bulls eye rash Headache Meningismus Arthralgias Pericarditis, usually a result of thoracic disease

Elevated cold agglutinin titers ( > 1:64) Mycoplasma IgM titers Sputum culture viral require special media

DFA staining of sputum Isolation of organism from sputum > 4 rise in titer between acute and convalescent sera Initial serum Legionella antibody titer ( > 1:256) Legionella urinary antigen Culture of pharyngeal swab C. pneumoniae IgM titer Serologic testing (western blot)

Presence of sulfur granules in specimen taken from sterile site (Continued )

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

273

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes

Nocardio spp.

Myocabacterium tuberculosis

Pericarditis Fungal Histoplasm capsulatum

Coccidioides immitis

Cryptococcus neoformans

Associated clinical features

Diagnostic tests

Chest pain Fever Weight loss Mass lesion or multiple small Cavities on chest X-ray Spontaneous drainage of empyema through chest wall Anorexia, weight loss Cough Dyspnea Hemoptysis Chest X-ray findings heterogeneous Often contiguous infection from lung Chest pain Weight loss Night sweats Dyspnea Cough Echocardiography may show multiple loculations of pericardial fluid Fever Weight loss Nonproductive cough Hilar adenopathy Patchy infiltrates on chest X-ray Pericarditis may result as an inflammatory response to inflamed mediastinal lymph nodes Fever Cough Night sweats Erythema nodosum Cardiac involvement rare

Organism cultured from tissue or pus Isolation of organism from pericardial fluid

Isolation of organism from blood, pus or sputum Isolation of organism from pericardium

Pericardial fluid usually shows lymphocytic predominance with elevated protein and moderately decreased glucose; pH usually 7.0–7.3 AFB smear of pericardial fluid rarely positive AFB culture of pericardial fluid

Blood pericardial fluid in acute primary pulmonary histoplasmosis Isolation of organism from sputum, blood cultures from pericardial fluid rarely positive

Isolation of organism from sputum (hazardous) Coccidioides immities IgM antibody from serum (high serum titers in disseminated disease) Headache, dizziness, somnolence, Culture of organism from CSF, sputum, urine, blood impaired memory, judgment India ink preparation of CSP may with CNS disease assist in presumptive diagnosis Dull chest pain, cough, dyspnea with pulmonary disease Skin lesions-papular, pustular, Visualization of organism in or nodular histological specimen Detection of cryptococcal antigen from serum and/or CSF (Continued )

274

Lazar et al.

Table 1 Causes of Myocarditis and Pericarditis, Clinical Features, and Diagnostic Tests (Continued ) Causes Aspergillus spp.

Candida spp.

Parasites Entamoeba histolytica

Toxoplasma gondii Occasionally

Associated clinical features

Diagnostic tests

High fever Isolation of organism from blood, Consolidation on chest X-ray CSF, bone marrow (rarely Rarely cerebral infarction due to positive) CNS invasion Isolation from sputum in Necrotizing skin lesions appropriate clinical setting Visualization of organism in histological specimen Fever Isolation of organism from blood Fluffy infiltrates in retina Visualization of organism in Nodular skin lesions histological specimen Pericarditis occurs as a Isolation of organism from stool complication of liver abscess Isolation of organism from liver Fever abscesses Weight loss Isolation of organism from Abdominal pain pericordial fluid Hepatic tenderness Elevated serum antiamebic (HI) Diarrhea antibody test Chest pain Asymtomatic cervical T. gondii IgM titer lymphadenopathy mononucleosis-like illness Toxoplasma cysts in tissue from biopsy specimen.

Abbreviations: CMV, cytomegalovirus; CNS, central nervous system, ELISA, enzyme-linked immunosorbent assay; IgG, immunoglobulinG; PCR, polymerase chain reaction; ESR, erythrocyte sedimentation rate; CSF, cerebrospinal fluid; PMN, polymorphonuclear leukocytes; AFB, acid-fast bacilli; CF, complement filtration; EIA, enzyme immunoassay; IFA, immunofluorescence assay; HIA, hemagglutination inhibition assay; LFT, liver function test; CSF, cerebrospinal fluid; LCM, lymphocytic choriomeningitis; ASO, antistreptolysin O.

occurs in approximately one-third of patients and is likely related to contiguous involvement of the pericardium (myopericarditis). Other physical findings include tachycardia disproportionate to the degree of fever, tachypnea, and narrowed pulse pressure from low cardiac output. LV failure results in leftward displacement of the apical impulse, the presence of a third and/or fourth heart sound, a systolic murmur of mitral regurgitation, and rales on pulmonary auscultation. Elevation of jugular venous pressure and peripheral edema arise from right-sided failure. The presence of a pericardial rub indicates contiguous pericarditis. This finding along with chest pain is important as pericardial involvement is predictive of future recurrences of myocarditis (15). Nonspecific markers of inflammation, including erythrocyte sedimentation rate and white blood cell count, are neither sensitive nor specific for myocarditis. In the Myocarditis Treatment Trial for acute and subacute forms of myocarditis, leukocytosis was present in one of five patients, and about half had elevated erythrocyte sedimentation rates (16). Increased white blood cell count and other markers of stronger immune response were associated with more profound cardiac dysfunction presentation. CPK-MB isoenzyme, a more specific marker of myocyte injury, is quite insensitive in active myocarditis, as only 6% of patients have elevated serum levels (17).

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

275

However, the frequency of CPK-MB elevation is higher (approximately 75%) in patients with ST segment elevation on electrocardiogram (EKG). By comparison, one-third of patients had elevated serum levels of Troponin I, another biochemical marker of myocardial injury (17). Troponin I did not correlate with the histological severity of myocarditis, but elevated serum levels were associated with symptoms of less than one-month duration. Serum brain natriuretic peptide levels, which provide diagnostic and prognostic value in patients with congestive heart failure, are likely nonspecific in myocarditis, but have not been well studied. The EKG is almost always abnormal in active myocarditis (14,18). ST-segment and T-wave changes are most common and may be diffuse or focal. The ST segments may be depressed or elevated. ST elevations are associated with elevation of CPK levels thereby prompting suspicion of recent MI. The presence of ST elevation without reciprocal ST depression has been proposed to be useful in differentiating myocarditis (19). The presence of pathological Q waves may further mimic MI. Other EKG findings include conduction abnormalities such as atrioventricular block and bundle branch block. Overall, EKG findings are rather nonspecific in suspected myocarditis. The chest X-ray may reveal cardiomegaly and pulmonary edema, although myocardial dysfunction may not be apparent on portable films that are taken in the critical care setting. Cardiac imaging modalities such as echocardiography and cardiac blood pool imaging may show both diffuse and regional wall motion abnormalities during active myocarditis. In one study, LV dysfunction was found in 69% of patients with biopsyproven myocarditis and in 88% of patients presenting with congestive heart failure (20). In most cases of LV dysfunction, the LV was normal in size and not dilated as expected in other types of cardiomyopathy. Failure of the LV to dilate might be related to decreased ventricular compliance associated with LV hypertrophy, a common echocardiographic and pathological finding in myocarditis, Nevertheless, reduced LV systolic function with normal LV chamber size in the setting of congestive heart failure provides another clue as to the possibility of myocarditis (21). LV dysfunction occurs in onefourth of patients with the right ventricular also normal in size in most cases. Magnetic resonance imaging using T2-weighted images detecting tissue water content as an indicator of inflammation may be useful. The addition of early and late gadolinium enhancement may help distinguish myocarditis from MI (22). Segmental early subendocardial defects with corresponding segmental subendocardial or transmural delayed high enhancement is characteristic of patients with MI, whereas patients with myocarditis exhibit normal first pass perfusion imaging, nonsegmental nonsubendocardial delayed enhancement (focal or diffuse) predominantly in the inferolateral location, and visualization of hyper-enhancing nodules. Although the diagnosis of myocarditis may be made with reasonable certainty on clinical grounds, endomyocardial biopsy remains the gold standard. The presence of myocyte injury and lymphocytic infiltrate are required for histological criteria. Recently, the role of endomyocardial biopsy has been questioned. In a randomized, placebo-controlled trial of 111 patients with LVEF, less than 45% had histological evidence of myocarditis immunology. Suppressive therapy consisting of prednisone with either cyclosporine or azathioprine did not affect survival or improve LVEF (16). The low incidence of myocarditis found on biopsy as well as the lack of demonstrable clinical benefit of immunosuppression has resulted in the utility of endomyocardial biopsy being challenged. Therefore, the utility of endomyocardial biopsy remains uncertain and the procedure is not routinely performed.

276

Lazar et al.

Noninvasive detection of myocardial inflammation has been attempted with gallium-67 imaging and indium-111 antimyosim antibodies, which bind to areas of myocardial necrosis. However, these techniques have not gained widespread acceptance for use in suspected myocarditis. Therefore acute myocarditis has a wide spectrum. LV dysfunction is central to the diagnosis but may have other causes such as sepsis-induced cardiac dysfunction. Recently, Tako-tsubo–like LV dysfunction has been described and is being recognized with increasing frequency. It is characterized by transient LV apical ballooning in the absence of coronary disease, associated with chest pain, dyspnea, and syncope. It is more common in females and occurs frequently after emotional or physical stress. The pathophysiology is poorly understood, but myocarditis had been hypothesized. The course of congestive heart failure related to myocarditis is quite variable. Myocarditis may resolve spontaneously with complete resolution of symptoms. Spontaneous improvement in LVEF is common. LV function returns to normal in about half of the patients, making it difficult to draw meaningful conclusions about the clinical effects of immunosuppression in uncontrolled trials. In the Myocarditis Treatment Trial, the mean LVEF improved from 26% to 34%, but as aforementioned, was unaffected by immunosuppressive agents (16). Higher LVEF and less intensive conventional therapy at baseline were independent predictors of survival. Patients with fulminant lymphocytic myocarditis have a better prognosis than those with acute nonfulminant myocarditis (23). CMV is the most common specific finding in immunocompetent patients with fatal myocarditis (24). The medical management of myocarditis heart failure includes diuretics and angiotensin-converting inhibitors. Digitalis may be used with caution as patients with myocarditis are particularly sensitive to its effects (2). Intravenous gamma globulin has been advocated to attenuate inflammatory cytokines during the acute treatment of myocarditis. Although a recent review identified three case series having shown gamma globulin to improve LV function, one randomized controlled trial of 62 patients found no benefit with respect to cardiac function, outcome, or event-free survival (25). Additional attempts at suppressing inflammation with aspirin and anti-inflammatory agents have been disappointing. In five animal studies of coxsacchie B3– and B4–induced myocarditis, aspirin, indomethacin, and ibuprofen resulted in a two to threefold increase in inflammation, myocyte necrosis, and mortality when compared to placebo (26). The deleterious effect was more prominent during the acute and subacute phases of myocarditis. The possible contribution of microvascular spasm to the progression of myocarditis toward dilated cardiomyopathy provides a rationale for the use of calcium channel blockers in acute myocarditis (2). However, their use in myocarditis has not been studied and should not be considered as multiple studies have demonstrated clinical deterioration associated with calcium channel blockers in congestive heart failure. Similarly, enhanced nuclear factor-kappaB expression suggests a protective effect of PPAR-gamma activators, which has not yet been studied. Bed rest is generally recommended to patients with active myocarditis as animal studies have shown myocardial necrosis to be enhanced by exercise (15). Critically ill patients often require mechanical circulatory support such as percutaneous. In one study, 71% of patients with fulminant myocarditis supported with percutaneous extracorporeal membrane oxygenation survived with spontaneous improvement of LV function (27). LV assist devices have been implanted in patients with myocarditis as a bridge to cardiac transplantation but has also recently been found to provide a bridge to

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

277

LV recovery in 11% of patients. Patients with myocarditis and postpartum cardiomyopathy are most likely to spontaneously recover (28). Myocardial Infarction Acute myocarditis should also be considered in suspected acute MI with normal coronary arteries and no other identifiable causes (21). Chest pain, fever, EKG changes, CPK-MB elevation, and regional wall motion abnormalities on cardiac imaging studies are common on both entities. As above, a young patient age, preceding viral syndrome, and ST-segment elevation on EKG in the absence of reciprocal ST depressions suggest acute myocarditis. Nonetheless, most cases of acute coronary syndrome result from atherosclerotic plaque rupture and thrombosis. Myocarditis is usually first suspected when normal coronary arteries are found on cardiac catheterization after patients are treated in a conventional manner with antianginal agents, anticoagulation, and thrombolytic therapy. Active myocarditis was found on biopsy in 33% of such patients up to six years after biopsy (29). Despite the theoretical concerns of myocardial hemorrhage and cardiac tamponade, one study found thrombolytic therapy to be uncomplicated in patients with viral myopericarditis misdiagnoses for acute MI (30). Although small in size, this study also showed a favorable long-term outcome for acute myocarditis mimicking acute MI without recurrent cardiac events. Arrhythmias Acute infectious myocarditis may also present with arrhythmias and conduction disturbances. Premature ventricular and supraventricular complexes occur most frequently and have been observed during the first three days of hospitalization in one-third of patients admitted for myocarditis (31). The overall incidence of arrhythmias likely decreases with time after presentation. In one study, complex ventricular ectopy decreased from 28% at one week to 8% at three months (31). Ventricular tachycardia is presumably less common, but may be responsible for 5% to 15% of cases of sudden death in young athletes related to myocarditis. The overall risk of malignant arrhythmias is unknown as its prevalence remains undetermined. One study found myocarditis to be present in four to six young asymptomatic athletes with minor rhythm disturbances and/or echocardiographic abnormalities (32). Ventricular arrhythmias and sudden cardiac death may be the initial manifestation of myocarditis and occur in the absence of LV dysfunction. Surviving patients should be referred for electrophysiological testing.

ACUTE PERICARDITIS Acute pericarditis is characterized clinically by chest pain, a pericardial friction rub, and serial electrocardiographic findings. The most commonly identified causes are infection, uremia, bacteria, acute MI, and trauma. Infectious Causes Many of the same agents that are responsible for the development of acute infectious myocarditis can cause pericarditis as well. In fact, illness most often manifests as a myopericarditis. The enteroviruses, especially Coxsackie group B (which is the agent of epidemic pleurodynia), are again the most commonly implicated pathogens (33).

278

Lazar et al.

Other viruses, such as CMV, EBV, mumps, herpes, varicella, and hepatitis B, have all been reported to cause pericarditis, but incidence is low (34). At one time, bacteria were the major cause of pericarditis, but since the development of antimicrobials, purulent pericarditis is unusual. It is occasionally encountered as a complication of pneumococcal and staphylococcal pneumonia or bacteremia. Hemophilus influenza type B was also responsible for the development of pericarditis in children, but the incidence of this is decreasing as well due to the use of the H. influenzae type B conjugate vaccine. Streptococcal and meningococcal infections may result in the development of pericardial disease, as can infection with L. pneumophila and M. pneumoniae. The aerobic gram-negative rods have been reported to cause disease, especially in debilitated or immunocompromised hosts (35,36). Mycobacterium tuberculosis is a well-known cause of primary pericarditis and results from hematogenous spread or direct extension of the infection to the pericardium. Cases that develop acutely may result in cardiac tamponade. The incidence of tuberculosis pericarditis has been rising, due to the increasing incidence of tuberculosis in the area of HIV infection (35,37). Pericarditis occurs with some disseminated fungal infections, especially histoplasmosis, aspergillosis, and candidiasis; however, it is unusual with coccidioidomycosis. Pericardial involvement is more commonly seen in immunocompromised individuals. Parasitic disease of the pericardium is rare, but cases of pericarditis secondary to T. gondii, E. histolytica, and schistosomes have been reported (35). The infectious causes of pericarditis and their associated clinical and laboratory features are listed in Table 1. Clinical Presentation The chest pain of pericarditis is classically described as a sharp or stabbing sensation located in the substernal or pericardial areas, with radiation to the left trapezius region (38). It is usually persistent, variable in intensity, and worsened by lying supine and relieved upon sitting up. Pleural-like pain may also be present. A pericardial cause should always be considered in the evaluation of pain in the trapezius ridge. Dyspnea is also a common symptom of acute pericarditis and is likely related to shallow respiration to avoid pleuropericardial discomfort. On physical examination, the most characteristic sign is a pericardial friction rub. Although the friction rub is pathognomonic for pericarditis, it is appreciated in only two-thirds of patients. Therefore, the absence of a rub does not exclude the diagnosis of pericarditis (38). The rub is a scratchy, superficial sound heard best at the lower left sternal border with the patient leaning forward using the diaphragm of the stethoscope. It often becomes obvious during inspiration. The rub typically has three components representing pericardial–epicardial contact during rapid ventricular filling, atrial systole, and ventricular systole. Although rubs may be confused with murmurs, the left parasternal location and the failure of a sound to radiate to areas expected of murmurs may help to identify a pericardial rub (39). The second diastolic component of atrial contraction is the most specific finding to pericarditis. Other physical findings include fever and tachycardia. Chest X ray may reveal enlargement of the cardiac silhouette due to pericardial effusion. Pleural effusions are evident in one-fourth of patients with acute pericarditis and are usually left-sided in contradistinction to the right-sided effusion seen in congestive heart failure (40). The presence of a pulmonary infiltrate of pleural effusion favors pleuroperocarditis or pulmonary infarction.

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

279

The EKG is abnormal in over 90% of patients (41). The classic evolutionary changes occur in half of the patients. During the first few hours of acute pericarditis, patients develop diffuse ST-segment elevations without reciprocal changes. ST depressions can occur in leads, AVR and V1. Of note, the T waves remain upright. Over weeks, the ST segments return to baseline and depression of the PR segment develops in 80% of patients (42). After the ST segments return to baseline, the T waves become inverted and ultimately normalized. Although both acute pericarditis and MI are associated with ST-segment elevation, several important EKG features may help to differentiate these two entities (41). ST elevations are diffuse in pericarditis and usually localized in acute MI. Pericarditis is more commonly associated with PR-segment depression and less commonly with Q waves than acute MI. In addition, T-wave inversion is present during ST-segment elevation in MI. However, in acute pericarditis, the T waves invert after the ST segments return to baseline. In other words, ‘‘the T waves flip after the STs dip.’’ Diffuse ST elevations may be seen in young healthy subjects with ‘‘early repolarization.’’ As compared to early repolarization, acute pericarditis is more likely when inspection of level V6 shows the height of the ST segment to exceed one-fourth the amplitude of the T wave (43). Low voltage may also be present in the presence of pericardial effusion. Of note, typical EKG findings occur less commonly in uremic pericarditis. Nonspecific markers of inflammation, including erythrocyte sedimentation rate and white blood cell count, are neither sensitive nor specific for pericarditis. Serum CPK levels are usually within normal limits but may be mild to moderately elevated with underlying myocarditis. Elevated CPK levels are associated with more prominent EKG changes. Patients clinically suspected of acute pericarditis should undergo echocardiography. Echocardiography is the most sensitive technique for detecting pericardial effusion. As pericardial fluid appears as an echo-free space, its size, distribution, and hemodynamic significance can be assessed (44). However, the absence of a pericardial effusion does not rule out the diagnosis of pericarditis, although its presence is further suggestive of the diagnosis. After exclusion of other ominous underlying conditions such as MI, the clinical course of acute pericarditis is usually self-limited. However, 10% to 20% of patients may develop chronic relapsing pericarditis (42). The management of acute pericarditis includes the symptomatic relief of chest pain with nonsteroidal anti-inflammatory agents and analgesics. Most patients should be admitted to the hospital for observation of complications including cardiac tamponade and arrhythmias. Aspirin and indomethacin are the initial agents used. The use of glucocorticoids is controversial as recurrences are common when therapy is discontinued. The complications of acute infectious pericarditis include cardiac tamponade and cardiac arrhythmias. Cardiac tamponade is the most frequent and often lifethreatening complication of acute pericarditis. Pericardial inflammation results in an effusion within the pericardial space that is detectable in half of the patients with pericarditis. Cardiac tamponade develops in 15% of patients with pericarditis and effusion. The hemodynamic significance of a pericardial effusion is more closely related to its rapidity of accumulation than to its volume. Rapid accumulation of an effusion does not allow for stretching of the pericardium. Consequently, tamponade may develop even in the presence of a small effusion, and large chronic effusions may not cause tamponade. Cardiac tamponade is not an all-or-none phenomenon, but is a spectrum of compressive physiology (45). Pericardial fluid restricts diastolic filling of the right ventricle and then the left ventricle, resulting in low cardiac output.

280

Lazar et al.

On physical examination, there is tachycardia, hypotension, and narrowed plus pressure reflecting decreased stroke volume. Pulsus paradoxus, an inspiratory decrease of >10 mmHg in systolic blood pressure during quiet respiration, is a late finding. In the absence of dehydration, there should be elevation of jugular venous pressure as systemic venous pressure is increased. On auscultation, the lungs are clear. Pulses alternans in which the amplitude of QRS complexes alternate from beat to beat is present. Echocardiographic findings include right atrial and right ventricular chamber collapse, inferior vena cava distension, and variation of mitral and tricuspid flow velocity with respiration on Doppler examination. Pericardiocentesis or open surgical drainage of the effusion is indicated for signs of cardiac tamponade. Although needle pericardiocentesis can usually be performed earlier, surgical drainage is considered safer. In the absence of tamponade, pericardiocentesis for laboratory evaluation of pericardial fluid carries a low diagnostic yield of 14% and is generally not recommended (46). Therapeutic pericardiocentesis provides a diagnostic etiology in about a third of cases.

Summary In summary, acute infectious myocarditis has a variable clinical course and a wide clinical spectrum ranging from asymptomatic EKG changes to progressive heart failure and death. It may also present with arrhythmia and/or sudden cardiac death or can mimic MI. The clinical factors that should raise suspicion of active myocarditis include young patient age, absence of cardiac history, recent or present viral syndrome, a brief duration of symptoms, coexisting pericarditis, and the absence of reciprocal ST depressions in suspected MI. Laboratory markers of active inflammation are generally not helpful. Cardiac imaging studies often reveal regional and diffuse wall motion abnormalities without chamber enlargement. Endomyocardial biopsy is required for definitive diagnosis; however, a strong presumptive diagnosis can be made clinically. In the appropriate clinical setting, a fourfold rise in acute and convalescent antibody titers, IgM-specific antibodies, isolation of virus from another site such as the throat or stool, or positive blood culture would strongly suggest the etiological agent of infection. Treatment of acute myocarditis is generally supportive. Acute pericarditis is characterized clinically by chest pain, a pericardial friction rub, and serial electrocardiographic changes. Although the presence of a pericardial friction rub is path gnomonic of pericarditis, its absence does not exclude the diagnosis. Differentiating electrocardiographic features from acute MI includes ST segment of diffuse nature, resolution of ST elevation prior to T-wave inversion, and PR-segment depression. Laboratory markers of active inflammation are generally not helpful. All patients suspected of pericarditis should be referred for echocardiography. Further management includes the treatment of chest pain with anti-inflammatory agents and analgesics as well as observation for complications, including cardiac tamponade and arrhythmias.

REFERENCES 1. Mason JW. Distinct forms of myocarditis. Circulation 1991; 83:1110–1111. 2. Sole MJ, Liu P. Viral myocarditis: a paradigm for understanding the pathogenesis and treatment of dilated cardiomyopathy. J Am Cull Cardiol 1993; 22:A99–A105.

Acute Myocarditis and Acute Pericarditis in the Critical Care Unit

281

3. Woodruff JF. Viral myocarditis: a review. Am J Pathol 1980; 101:427–478. 4. See DM, Tilles JG. Viral myocarditis. Rev Infec Dis 1991; 13:951–956. 5. Savoia MC, Oxman MN. Myocarditis, pericarditis. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases, 4th ed. New York: Churchill Livingstone, 1995:799–813. 6. Maje SS, Adolph RJ. Myocarditis; unresolved issues in diagnosis and treatment. Clin Cardiol 1990; 13:69–79. 7. Obeyeshere I, Hermon Y. Myocarditis and cardiomyopathy after arbovirus infections. Br Heart J 1972; 34:821–827. 8. Gore J. Myocardial changes in fatal diphtheria: summary of observations in 221 cases. Am J Med Sci 1948; 215:257–266. 9. Smith RH, Lazar JM. Myocarditis. Infect Dis Prac 1996; 20:94–95. 10. Nontrowitz NE, Smith RH. Myocarditis update. Emerg Med 1993; 25:69–74. 11. Lerner AM. Pericarditis-myocarditis. In: Gorbach SL, Bartlett JG, Blacklow NR, eds. Infectious Diseases. Philadelphia: WS Saunders, 1992:565–572. 12. Atkinson JB, Connor DH, Robinowitz M, et al. Cardiac fungal infections. Review of autopsy findings in 60 patients. Hum Pathol 1984; 15:935–942. 13. Chow LC, Dittrich HC, Shabetai R. Endocardial biopsy in patients with unexplained congestive heart failure. Ann Intern Med 1988; 109:535–539. 14. Lewes D, Rainford DJ, Lane WF. Symptomless myocarditis and myalgia in viral and mycoplasma pneumonia infections. Br Heart J 1974; 36:924–932. 15. Johnson RA, Palacios I. Dilated cardiomyopathies of the adult. N Engl J Med 1982; 307:1119–1126. 16. Mason JW, O’Connell JB, Hershkowitz A, et al. A clinical trial of immunosuppressive therapy for myocarditis. N Eng J Med 1995; 333:269–275. 17. Smith SC, Ladenson JH, Mason JW, et al. Elevations of cardiac troponin I associated with myocarditis. Circulation 1997; 95:163–168. 18. Karjalainen J, Heikkila J. ‘‘Acute pericarditis’’: myocardial enzyme release as evidence for myocarditis. Am Heart J 1986; 111:546–552. 19. Nakashima H, Honda Y, Katayama T. Serial electrocardiographic findings in acute myocarditis. Intern Med 1994; 33:659–666. 20. Pinamonti B, Albert E, Cigalotto A, et al. Echocardiographic findings in myocarditis. Am Coll Cardiol 1992; 20:85–89. 21. Laissey JP, Hyafil F, Juliard JM, et al. Differentiating acute myocardial infarction from myocarditis: diagnostic value of early-and delayed-perfusion cardiac MR imaging. Radiology 2005; 237:75–82. 22. Dec GW Jr, Waldman H, Fallon JT, et al. Viral myocarditis mimicking acute myocardial infraction. J Am Coll Cardiol 1992; 20:85–89. 23. McCarthy RE III, Boehmer JP, Hruban RH, et al. Long-term outcome of fulminant myocarditis as compared with acute (non-fulminant) myocarditis. N Eng J Med 2000; 342:690–695. 24. Kyto V, Vuorinen, Saukko P, Lautenschlager I, et al. Cytomegalovirus infection of the heart is common in patients with fatal myocarditis. Clin Infect Dis 2005; 40:683–688. 25. Robinson JL, Hartling L, Crumley E, et al. A systematic review of intravenous gamma globulin for therapy of acute myocarditis. BMC Cardiovasc Disord 2005; 5:12–17. 26. Meune C, Spaulding C, Mahe I, et al. Risks versus benefits of NSAIDS including aspirin in myocarditis: a review of the evidence from animal studies. Drug Safety 2003; 26(13):975–981. 27. Asaumi Y, Yasuda S, Morii I, et al. Favourable clinical outcome in patients with cardiogenic shock due to fulminant myocarditis supported by percutaneous extracorporeal membrane oxygenation. Eur Heart J 2005; 26:2185–2192. 28. Simon MA, Kormos R, Murali S, et al. Myocardial recovery using ventricular assist devices. Circ 2005; 112(SI):I32–I36. 29. Dec GW, Palacios IF, Fallon JT, et al. Active myocarditis in the spectrum of acute dilated cardiomyopathies. N Eng J Med 1985; 312:885–890.

282

Lazar et al.

30. Millaire A, de Groote P, Decoulx E, et al. Outcome after thrombolytic therapy of nine cases of myopericarditis misdiagnosed as myocardial infarction. Eur Heart J 1995; 16:333–338. 31. Vikerfors T, Stjema A, Olcen P, et al. Acute myocarditis. Acute Med Scand 1988; 223:45–52. 32. Zeppolli P, Santini C, Palmieri V, et al. Role of myocarditis in athletes with minor arrhythmias and/or echocardiographic abnormalities. Chest 1994; 106:373–380. 33. Smith WG. Coxsakie B myopericarditis in adults. Am Heart J 1970; 80:34–36. 34. Fowler No, Manitas GT. Infectious pericarditis. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases. 4th ed. New York: Churchill Livingstone, 1955:799–815. 35. Savoia MC, Oxman MV. Myocarditis and pericarditis. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases. 4th ed. New York: Churchill Livingstone, 1995:799–815. 36. Maisch B. Pericardial diseases with a focus on etiology. Curr Opin Cardiol 1994; 9: 379–388. 37. Ortbals DW, Avioli LV. Tuberculous pericarditis. Arch Intern Med 1979; 139:231–234. 38. Shabetai R. Acute pericarditis. Cardiol Clin 1990; 8:639–671. 39. Spodick DH. Diagnostic electrocardiographic sequences in acute pericarditis: significance of PR segment and PR vector changes. Circulation 1973; 48:575. 40. Weiss JM, Spodick DH. Association of left pleural effusion with pericardial disease. N Engl J Med 1983; 308:696. 41. Spodick DH. Differential diagnosis of acute pericarditis. Prog Cardiovasc Dis 1971; 14:192. 42. Spodick DH. Diagnostic electrocardiogram in acute pericarditis: distributions of morphologic and axial changes by stages. Am J Cardiol 1974; 33:470. 43. Ginzton LE, Laks MM. The differential diagnosis of acute pericarditis from the normal variant: new electrocardiographic criteria. Circulation 1982; 65:1004. 44. Kronzon I, Cohen ML, Winer HE. Contribution of echocardiography to the understanding of the pathophysiology of cardiac tamponade. J Am Coll Cardiol 1983; 1:1180–1182. 45. Reddy PS, Curtiss EI. Cardiac tamponade. Cardiol Clin 1990; 8:627–633. 46. Permanyer-Miralda G, Sagrista-Sauleda J, Soler-Soler J. Primary acute pericardial disease. A prospective series of 231 consecutive patients. Am J Cardiol 1985; 56:623.

13 Central Intravenous Line Infections in the Critical Care Unit Burke A. Cunha Infectious Disease Division, Winthrop-University Hospital, Mineola, and State University of New York School of Medicine, Stony Brook, New York, U.S.A.

INTRODUCTION Intravenous (IV) central venous catheters (CVC) are used in critical care units (CCU) for medication, fluid, or nutritional access. IV CVCs may be inserted peripherally, i.e., peripherally inserted central catheters in central veins. Complications of CVCs may be mechanical/infectious. The three most common infectious complications of CVC include bacteremia, septic thrombophlebitis, and acute bacterial endocarditis (ABE). The most common organisms associated with CVC infection are methicillinsensitive Staphylococcus aureus (MSSA)/methicillin-resistant S. aureus (MRSA), S. epidermitis, also known as coagulase-negative staphylococci (CoNS), and less commonly aerobic gram-negative bacilli. Fungal IV CVC infections may occur in any patient with CVCs in place for an extended period of time or receiving total parental nutrition. Enterococci are uncommon causes of CVC, excluding femoral lines. Because most patients in CCUs have one or more CVCs, clinicians caring for patients in the CCU should be familiar with the infectious complications of CVC. Physicians consulting in the CCU should be familiar with the differential diagnosis therapy of CVC infections (1–10).

OVERVIEW OF CVC INFECTIONS There are several factors that predisposed to CVC infections. After aseptic insertion technique, the most important factors predisposing to infection are duration and location of insertion of CVCs. IV CVC line infections are a function of time. CVC-related line infection is rare before seven days; or after seven days, there is a gradual increase over time in the incidence of IV CVC line infections. The number of lumens may increase the potential for IV CVC infection. In a patient with otherwise unexplained fever in the CCU, the longer a CVC is in place the more likely the CVC is the likely cause of fever. Other important determinants of IV CVC line 283

284

Cunha

Table 1 Pathogens Associated with Intravenous-Line Infections Most common pathogens Staphylococcus aureus (MSSA/MRSA) Staphylococcus epidermidis/CoNS Enterobacter agglomerans Enterobacter cloacae Less common microorganisms Enterococci (VSE/VRE; vancomysin-sensitive enterococci/ vancomycin-resistant enterococci) Burkholderia (Pseudomonas) cepacia Stenotrophomonas (Xanthomonas) maltophilia Citrobacter freundii Abbreviations: MSSA, methicillin-sensitive S. aureus; MRSA, methicillinresistant S. aureus; CoNS, coagulase-negative staphylococci. Source: From Ref. 1.

infections are the anatomical location of CVC. The best anatomical location with the lowest potential for infection is the subclavian vein. The next best location is the internal jugular vein, and the least desirable location from an infectious perspective is the femoral location. It should be appreciated that peripheral IV lines rarely result in IV line infections. Even if phlebitis and bacteremia develops from a peripheral IV line, the discontinuous/low-grade bacteremia does not result in endocarditis. The reticuloendothelial system rapidly eliminates microorganisms introduced in the blood stream via peripheral IV lines. If peripheral IV lines are associated with a intermittent/low-grade S. aureus bacteremia, S. aureus ABE is not a complication, (Tables 1 and 2) (1–5,11–18). IV LINE INFECTIONS The main diagnostic difficulty with CVC infections is that only 50% of CVCs that are infected have any indicators of infection present locally. CVC IV line infection is straightforward when the insertion site is red and painful. But half of the time there are no superficial signs of IV line infection. It is usually straightforward to differentiate chemical phlebitis or IV line infiltration from cellulitis at the CVC insertion site. Table 2 Risk Factors Associated with Central Intravenous-Line Infections Important risk factors for central IV line infections Aseptic insertion technique Duration of catheterization (catheter days) Location of catheter placement Multiple lines Less important factors in central IV line infections Contaminated infusate Number of catheter lumens (single vs. triple lumen) Secondary bacteremias Host defense status Abbreviation: IV, intravenous. Source: From Refs. 1, 3, 4.

Central Intravenous Line Infections in the Critical Care Unit

285

The skin at the IV insertion site in IV phlebitis or with IV infiltration is swollen and painful but is not erythematous and the patient does not have otherwise unexplained fever. IV line infections secondary to CVC must be suspected in CCU patients with fever where the other causes of fever in the CCU have been ruled out. Then the diagnosis of CVC-related infection should be entertained. The likelihood of CVC-related infection is more likely the longer the CVC has been in place and, as mentioned, is also related to the anatomical location of the insertion, i.e., femoral lines are much more likely to become infected than subclavian lines (1–8). IV line infection in the absence of local manifestations may be diagnosed by blood cultures and semiquantitative catheter tip cultures. If IV line infection is suspected from a CVC, the catheter should be removed and the tip sent for a semiquantitative culture. Simultaneously, the patient should have blood cultures drawn, but not drawn through the removed or other CVCs unless there is no venous access. IV line infection is diagnosed if the blood culture isolate, excluding skin contaminants acquired during venipuncture, is the same organism recovered from the removed CVC tip culture. For the CVC tip culture to be considered positive, 15 colonies should be present. Positive tip cultures without bacteremia indicate the catheter colonization of noninfection. Bacteremia without a positive CVC tip culture indicates bacteremia unrelated to the line. IV line infection is only diagnosed if there are 15 or greater colonies, from culturing the removed catheter tip, and they are the same species as the blood culture isolates (1,3,15–19). The therapy of uncomplicated IV line infections is for two weeks with antiMSSA/MSRA antibiotics. Near the end of the therapy, MSSA/MRSA ABE should be ruled out by transthoracic echocardiography (TTE)/transesophageal echocardiography (TEE). Elevated teichoic acid antibody titers/otherwise unexplained elevated erythrocytes sedimentation rate (ESR) is a clue to the possibility of ABE following MSSA/MRSA bacteremia. Cardiac echocardiography should be done in a patient with a high-grade/persistent S. aureus bacteremia following an IV CVC infection, particularly with an otherwise unexplained elevated ESR or teichoic acid antibody level. In patients without prosthetic valves. TTE is sufficient; TEE is a low-yield procedure. For prosthetic valves, TEE is preferred. Antimicrobial coated CVCs have not been shown to decrease IV line infection greater than seven days after placement (20–35).

SEPTIC THROMBOPHLEBITIS Simple uncomplicated phlebitis may be associated with low-grade fevers 102 F and is not associated with bacteremia. If bacteremia complicates phlebitis, it is due to skin organisms, usually S. aureus or CoNS, and the bacteremia is intermittent and is of low intensity. Typically, if blood cultures are positive, they are present in a 1/2, 2/0, 2/2, 0/4, etc., indicative of blood culture contaminants. Septic thrombophlebitis is an IV septic process within the vein. The clinical findings resemble phlebitis except that the patient usually has fevers of 102 F and is often accompanied by rigors. Blood culture positivity may be continuous or discontinuous and is usually of a high-grade 3/4 or 4/4 positive blood cultures. The diagnosis of septic thrombophlebitis may be suspected clinically and confirmed by removal of the central IV line, and pus emanating from the catheter wound and palpable cord is usually present. Therapy for septic thrombophlebitis is venotomy, with appropriate anti MSSA/MRSA therapy for four weeks.

286

Cunha

S. AUREUS ACUTE BACTERIAL ENDOCARDITIS S. aureus is the commonest cause of ABE at the present time. S. aureus, either MSSA/MRSA, is capable of attacking normal and abnormal heart valves. This is in contrast to SBE due to avirulent pathogens, e.g., Staphylococcus veridans group that requires pre-existing valvular damage and capsular production to cause SBE. The factors that predispose to MSSA/MRSA ABE include high-grade/continuous MSSA/MRSA bacteremia from a CVC, invasive cardiac procedure, e.g., radiofrequency ablation or a distant protected focus, e.g., abscess. ABEis not a complication of peripheral IV line infection. The clinical diagnosis of S. aureus ABE demonstrates two diagnostic features. First, the patient must have a high-grade/continuous S. aureus bacteremia, i.e., 3/4 or 4/4 positive blood cultures. The second criterion is the demonstration of vegetation by TTE/TEE. S. aureus bacteremia that is not high grade/prolonged indicates a transient staphylococcal bacteremia and is not indicative of endocarditis per se. In S. aureus endocarditis, the bacteremia characteristically is of high grade and prolonged. High-grade prolonged S. aureus bacteremia without a vegetation demonstrated by TTE/TEE should suggest an extracardiac focus or abscess. Patients with ABE may or may not have a new/rapidly changing cardiac murmur. If the endocarditis is recent, there may have been sufficient valvular damage to result in a cardiac murmur. There is no indication to get a TTE/TEE to rule out endocarditis in patients without bacteremia. If a cardiac vegetation is demonstrated and no concomitant bacteremia is present from an organism that is a known endocarditis pathogen, then the vegetation is an incidental finding and not indicative of ABE. Sterile vegetations, also known as marantic endocarditis, may occur in association with malignancy as well as nonmalignant disorders, e.g., Liebman–Saks endocarditis. Therefore the diagnosis of S. aureus ABE rests on demonstrating a high-grade continuous bacteremia that is persistent in a patient with demonstrable vegetation by cardiac EKG, murmur may or may not be present. In non-IVDAs, the fever in ABE is usually 102 F (Table 3). The treatment of MSSA/MRSA ABE is for four to six weeks of anti-staphylococcal therapy. For MSSA ABE, treatment is usually with oxacillin, nafcillin, and first-generation cephalosporins, e.g., cephazolin. In penicillin-allergic patients with MSSA/MRSA ABE, vancomycin quinupristin/dalfopristin, minocycline, linezolid, and daptomycin have been used. All of the drugs used to treat MRSA are also effective against MSSA, but the reverse is not true. Because the therapy of MRSA/MSSA is prolonged, i.e., four to six weeks, oral therapy for all or part of the therapy is desirable. The only two antibiotics available to treat MRSA ABE orally are minocycline and linezolid. If fever/bacteremia persists after a week of appropriate therapy then the clinician should re-evaluate drug treatment to be sure that the drug is being dosed optimally, as well as the nonantibiotic causes of apparent antibiotic failure, i.e., myocardial abscess and noncardiac septic foci. If the problem is drug-related, daptomycin offers the best option for terminating the bacteremia/curing the endocarditis. Daptomycin is concentration-dependent on the killing kinetics and is the most potent anti-staphylococcal antibiotic available against MSSA/MRSA. Although the usual dose of daptomycin for bacteremia/ABE is 6 mg/kg (IV) q24 hours (with normal renal function), the dose of daptomycin may be increased if the patient is not responding to daptomycin or other anti-staphylococcal antibiotics. Daptomycin given at a dose of 12 mg/kg (IV) q24 hours (with normal renal function) has been used safely without side effects for over four weeks (36–44). If the problem is not drug-related and is related to a

Central Intravenous Line Infections in the Critical Care Unit

287

Table 3 Infectious Complications of Central Venous Catheters Intravenous line–related bacteremias Diagnostic features Low grade/short duration/discontinuous MSSA/MRSA bacteremias Temperatures usually 102 F Therapy Remove CVC Antibiotic therapy  2 wks Septic thrombophlebitis Diagnostic features IV line infection Usually high grade/continuous bacteremias Pus from CVC site when removed  Palpable venous cord Temperatures usually 102 F TTE/TEE negative if no ABE Therapy Remove CVC Empiric antibiotic therapy  2–4 wks Venotomy MSSA/MRSA ABE Diagnostic features High grade/continuous bacteremia Cardiac vegetation by TTE/TEE Cardiac murmur may not be present early ESR " usually 30–50 mm/hr range Teichoic acid antibody titers usually elevated (>1:4) Therapy Antibiotic treatment directed against MRSA until susceptibility to methicillin known. Treat MSSA or MRSA, ABE for 4–6 wks Abbreviations: CVC, central venous catheters; IV, intravenous; TTE, transthoracic echocardiography; TEE, transesophageal echocardiography; MRSA, methicillin-resistant S. aureus; MSSA, methicillinsensitive S. aureus; ESR, erythrocytes sedimentation rate; ABE, acute bacterial endocarditis. Source: From Refs. 1, 39, 42.

protected focus, e.g., abscess myocardial, S. aureus, acute bacteremia meningitis, or extra cardiac septic complications, then surgical drainage may be needed to eradicate the infection. REFERENCES 1. Cunha BA. Intravenous line infections. Crit Care Clin 1998; 8:339–346. 2. Sherertz RJ. Update on vascular catheter infections. Curr Opin Infect Dis 2004; 17: 303–307. 3. Seifert H, Jansen B, Farr BM. Catheter-related infections. 2d ed. New York: Marcel Dekker, 2004. 4. Maki DG. Infections caused by intravascular devices used for infusion therapy: pathogenesis, prevention, and management. In: Infections Associated with Indwelling Medical Devices. 2d ed. Washington: American Society for Microbiology, 1994. 5. Norwood S, Ruby A, Civetta J, Cortes V. Catheter-related infections and associated septicemia. Chest 1991; 99:968–975.

288

Cunha

6. Johnson A, Oppenheim BA. Vascular catheter-related sepsis: diagnosis and prevention. J Hosp Infect 1992; 20:67–78. 7. Toltzis P, Goldman DA. Current issues in central venous catheter infection. Annu Rev Med 1990; 41:169–176. 8. Corona ML, Peters SG, Narr BJ, Thompson RL. Infections related to central venous catheters. Mayo Clin Proc 1990; 65:979–986. 9. Arnow PM, Quimosing EM, Beach M. Consequences of intravascular catheter sepsis. Clin Infect Dis 1993; 16:778–784. 10. Farr BM, Hanna H, Raad I. Nosocomial infections related to use of intravascular devices inserted for long term vascular access. In: Mayhall G, ed. Hospital Epidemiology and Infection Control. 3rd ed. Baltimore: Lippincott Williams & Wilkins, 2004. 11. Raad I, Bodey GP. Infectious complications of indwelling long-term central venous catheters. Infect Dis 1992; 15:197–210. 12. Clarke DE, Raffin. Infectious complications of indwelling long-term central venous catheters. Chest 1990; 97:966–972. 13. Sitges-Serra A, Pi-Suner T, Garces JM, Segura M. Pathogenesis and prevention of catheter-related septicemia. Am J Infect Control 1995; 23:310–316. 14. Read II, Hohn DC, Gilbreath J, et al. Prevention of central venous catheter-related infections by using maximal sterile barrier precautions during insertion. Infect Control Hosp Epidemol 1994; 15:231–238. 15. Early TF, Gregory RT, Wheeler JR, Snyder SO, Gayle RG. Increased infection rate in double lumen versus single lumen Hickman catheters in cancer patients. S Med J 1990; 83:34–36. 16. Moro ML, Vigano EF, Lepri AC. The Central Venous Catheter-Related Infections Study Group. Risk factors for central venous catheter-related infections in surgical and intensive care units. Infect Control Hosp Epidemiol 1994; 15:253–264. 17. Ullman RF, Gurevich I, Schoch PE, Cunha BA. Colonization and bacteremia related to duration of triple-lumen intravascular catheter placement. Am J Infect Control 1990; 18: 201–207. 18. Raad I, Costerton W, Sabharwal U, Sacilowski M, Anaissie E, Bodey GP. Ultrastructural analysis of indwelling vascular catheters: a quantitative relationship between luminal colonization and duration of placement. J Infect Dis 1993; 168:400–407. 19. Maki DG, Weise CE, Sarafin HW. A semiquantitative culture method for identifying intravenous-catheter-related infection. N Engl J Med 1977; 296:1305–1309. 20. Hill PC, Birch M, Chambers S, et al. Prospective study of 424 cases of Staphylococcus aureus bacteremia: determination of factors affecting incidence and mortality. Intern Med J 2001; 31:97–107. 21. Fatkenheuer G, Preuss M, Salzberger B, et al. Long term outcome and quality of care patients with Staphylococcus aureus bacteremia. Eur J Clin Microbiol Infect Dis 2004; 23:157–162. 22. Kim SH, Park WB, Lee KD, et al. Outcome of Staphylococcus aureus bacteremia in patients with eradicable foci versus noneradicable foci. Clin Infect Dis 2003; 37:794–797. 23. Mylotte JM, McDermott C. Staphylococcus aureus bacteremia caused by infected intravenous catheters. Am J Infect Control 1987; 15:1–6. 24. Malanoski GJ, Samore MH, Pefanis A, Karchmer AW. Staphylococcus aureus catheterassociated bacteremia. Arch Intern Med 1995; 155:1161–1166. 25. Jensen AG. Importance of focus identification in the treatment of Staphylococcus aureus bacteremia. J Hosp Infect 2002; 52:29–36. 26. Finkelstein R, Sobel JD, Nagler A, Merzbach D. Staphylococcus aureus bacteremia endocarditis: comparison of nosocomial community-acquired infection. J Med 1984; 15: 193–211. 27. Lesens O, Hansmann Y, Storck D, Christmann D. Risk factors for metastatic infection in patients with Staphylococcus aureus bacteremia with and without endocarditis. Eur J Intern Med 2003; 14:227–231.

Central Intravenous Line Infections in the Critical Care Unit

289

28. Espersen F, Frimodt-Moller N. Staphylococcus aureus endocarditis: a review of 119 cases. Arch Intern Med 1986; 146:1118–1121. 29. Chang FY, MacDonald BB, Peacock JE, et al. A prospective multicenter study of Staphylococcus aureus bacteremia. Incidence of endocarditis, risk factors for mortality, and clinical impact of methicillin resistance. Medicine 2003; 82:322–332. 30. Bayer AS, Lam K, Ginzton L, Norman DC, Chiu CY, Ward JL. Staphylococcus aureus bacteremia: clinical serological echocardiographic findings in patients with and without endocarditis. Arch Intern Med 1987; 147:457–462. 31. Mirimanoff RO, Glauser MP. Endocarditis during Staphylococcus aureus septicemia in a population of non-drug addicts. Arch Intern Med 1982; 142:1311–1313. 32. Kaech C, Elzi L, Sendi P, et al. Course and outcome of Staphylococcus aureus bacteraemia: a retrospective analysis of 308 episodes in a Swiss tertiary-care centre. Clin Microbio Infect 2006; 12:345–352. 33. Fowler VG Jr., Olsen MK, Corey GR, et al. Clinical identifiers of complicated Staphylococcus aureus bacteremia. Arch Intern Med 2003; 163:2066–2072. 34. Raad II, Sabbagh MF. Optimal duration of therapy for catheter-related Staphylococcus aureus bacteremia: a study of 55 cases and review. Clin Infect Dis 1992; 14:75–82. 35. Rupp ME, Lisco SJ, Lipsett PA, et al. Effect of a second-generation venous catheter impregnated with chlorhexidine and silver sulfadiazine on central catheter-related infections—a randomized, controlled trial. Ann Inter Med 2005; 143:570–580. 36. Sacks-Berg A, Strampfer MJ, Cunha BA. Intravenous line sepsis due to suppurative thrombophlebitis. Heart Lung 1987; 16:318–320. 37. Mayhall CG. Diagnosis and management of infections of implantable devices used for prolonged venous access. Curr Clin Top Infect Dis 1992; 12:83–110. 38. Cunha BA. Clinical usefulness of highly elevated teichoic acid antibody (TAA) titers. Infect Disease Pract 2005; 29:378–380. 39. Cunha BA. Persistent S. aureus bacteremia: a clinical approach. Infect Disease Pract 2005; 29:444–446. 40. Cunha BA, Hamid N, Kessler, Parchuri S. Daptomycin cure after cefazolin treatment failure of Methicillin-sensitive Staphylococcus aureus (MSSA) tricuspid valve acute bacterial endocarditis from a peripherally inserted central catheter (PICC) line. Heart Lung 2005; 34:442–447. 41. Cunha BA, Eisenstein L, Hamid N. Pacemaker induced S. aureus mitral valve acute bacterial endocartitis complicated by persistent bacteremia from an infection coronary stent: cure with prolonged/high dose daptomycin without toxicity. Heart & Lung 2006; 35:217–222. 42. Cunha BA. Clinical manifestations and antimicrobial therapy of methicillin resistant Staphylococcus aureus (MRSA). Clinical Microbiology & Infection 2005; 11:33–42. 43. Cunha BA, ed. Antibiotic Essentials. 5th ed. Royal Oak, Michigan: Physicians Press, 2006.

14 Intra-Abdominal Surgical Infections and Their Mimics in the Critical Care Unit Meghann L. Kaiser and Samuel E. Wilson Department of Surgery, University of California, Irvine School of Medicine, Orange, California, U.S.A.

INTRODUCTION Postsurgical patients in the intensive care unit (ICU) often confront a myriad of medical and new surgical complications. Among these, intra-abdominal infections remain the most formidable adversary, affecting an estimated 6% of all critically ill surgical patients. Organ dysfunction continues to be a major manifestation of these infections, resulting in a high mortality of 23% (1). Intra-abdominal infection in the surgical ICU (SICU) patient may occur as a complication of a previous condition or arise de novo. In either event, it is evident that the critically ill patient is predisposed to a different set of disease states and pathogens than the clinician might routinely encounter. Moreover, given the complex background of concomitant illnesses in these individuals, physicians must be prepared to interpret a variety of atypical presentations. The burden of the diagnostician in the care of the ICU patient, however, remains not only of sensitivity but also of specificity; accordingly, the physician must be alert to a variety of clinical pictures that may masquerade as abdominal infection in the SICU patient. In this chapter, we will review the unique characteristics of intra-abdominal infections in critically ill patients, as well as the challenges faced in their diagnosis and treatment.

TERTIARY PERITONITIS With a startling mortality of 20% to 50%, the diagnosis and treatment of tertiary peritonitis has remained a source of intense research for some time (2). Tertiary peritonitis, or intra-abdominal infection persisting beyond a failed surgical attempt to eradicate secondary peritonitis, represents a blurring of the clinical continuum, often characterized by the lack of typically presenting signs and symptoms. Nevertheless, prompt diagnosis is essential for cure, and given the grim propensity of this complication to strike already critically ill patients—rapidly devolving into multi-organ

291

292

Kaiser and Wilson

system failure—the intensivist should be equipped with the necessary knowledge to suspect, confirm, and treat this serious illness. Early Recognition The gradual postoperative transitional period between a diagnosis of secondary and tertiary peritonitis causes the clinical presentation of tertiary peritonitis to be quite subtle. Moreover, because patients are frequently sedated, intubated, or otherwise incapacitated, history and physical exam in the early stages of disease are often an insensitive means to a diagnosis. Therefore, the physician must pay particular attention to those secondary peritonitis patients whose conditions place them at risk, including malnutrition and the several variables detailed under the acute physiological and chronic health evaluation score (APACHE) II scoring system such as age, chronic health conditions, and certain physiologic abnormalities while in the ICU (3). In these individuals, fever, elevated C reactive protein (CRP), and leukocytosis—although admittedly nonspecific in the postsurgical patient—should be addressed quickly and assertively, even when lacking other evidence of infection, such as abdominal tenderness and absent bowel sounds (3). As one might reasonably predict, clinical evidence of tertiary peritonitis becomes increasingly more obvious the farther the disease has progressed, eventually leading to multi-organ system failure. To this end, further scoring systems have been developed to determine the probability that tertiary peritonitis is in fact present postsurgically. Two such systems, the Sepsis-related Organ Failure Assessment and the Goris scores, attempt to objectively sum the failure of the respiratory, cardiovascular, nervous, renal, hepatic and coagulation systems. Even though first postoperative day scores are elevated in patients both with and without tertiary peritonitis, subsequent second and third day scores are seen to fall in those without the disease, whereas remaining steady in patients later diagnosed by re-operation with tertiary peritonitis (4). Although these findings may be interesting and statistically significant, their clinical application—in overall terms of mortality avoided—remains to be proven. By pausing for evidence of changing widespread system failure over time, the clinician risks losing the opportunity to avoid medical catastrophe. Radiologic tools, then, become a mainstay of the physician’s investigation. Two such studies, gallium-67 scintigraphy and computed tomography (CT) scan, are commonly used for the detection of intra-abdominal infection. On the whole, CT is generally the preferred choice. At 97.1% accuracy, it is the more accurate of the two, with an enviable specificity of 100%. Isotope scans suffer in terms of accuracy for the postoperative patient because of false-positive uptake in areas of surgical injury. Moreover, CT has the potential to contribute both diagnostically and therapeutically in the care of patients, as will be discussed later. Finally, CT may be done on demand, whereas Ga-67 requires one to two days for concentration of the isotope at the site of infection. Scintigraphy, however, is not entirely without its own merits. With a sensitivity of 100% relative to 93.7% for CT, it is superior for uncovering early infection prior to the development of discreet fluid collections. Also, it is worth considering that in centers where indium-111 and technitium-99m exametazinelabeled leukocyte scans are available, a higher level of scintigraphy accuracy may be attained, albeit at greater expense. Furthermore, as an incidental advantage, nucleotide scanning has been known to reveal extra-abdominal infections such as pneumonia and cellulitis that might imitate tertiary peritonitis (5). Therefore, one might consider this as a second option for the relatively stable patient, in which CT has failed to provide a definitive answer but signs and symptoms persist. Other studies, such as plain film and

Intra-Abdominal Surgical Infections and Their Mimics in the Critical Care Unit

293

ultrasound, are impaired by the nonspecific finding of intra-peritoneal free air and other features that might normally be expected in the postoperative patient (6). Microbiology and Pathogenesis The flora of tertiary peritonitis is different from that of secondary peritonitis. Whereas a culture of secondary peritonitis might produce a predominance of Escherichia coli, Streptococci, and Bacteroides—all normal gut flora—tertiary peritonitis is more apt to culture Pseudomonas, coagulase-negative Staphylococcus, Enterococci, and Candida (7,8). The obvious explanation for these differences is the mode of infection: secondary peritonitis is typically community acquired, but tertiary peritonitis occurs in an ICU setting. Time spent in the ICU necessarily implies that the patients affected are critically ill and likely already treated with antimicrobials. Some theorize that disease begins when the gut is weakened by surgical manipulation, hypoperfusion, antibiotic elimination of normal gut flora, and a lack of enteral feeding, thereby creating an opportunity for selected resistant native bacteria to translocate across the mucosal border (9). In fact, independent risk factors for postsurgical enterococcal infection include APACHE II scores greater than 12 and inadequate antibiotic coverage (8). Therefore, empiric antibiotic therapy should be broadly launched to cover the wide range of likely organisms, and later targeted to the specific determined pathogen and sensitivity. Appropriate first agents include, among others, carbapenems or the anti-pseudomonal penicillins, or a regimen of aminoglycosides with either clindamycin or metronidazole for the penicillin-allergic patient (6). Treatment When possible in selected patients, the treatment of tertiary peritonitis may be accomplished by image-guided percutaneous drainage of intra-abdominal abscesses, generally using CT. Percutaneous drainage is not without its inconveniences: complications such as fistulas, cellulitis, and obstructed, displaced, or prematurely removed drains occur in 20% to 40% of patients (10,11). Nevertheless, the efficacy of this technique is real: Cinat et al. found this method to be 90% successful in postoperative abscess. Abscesses involving the appendix, liver or biliary tract, and colon or rectum were also found to be particularly responsive at rates of 95%, 85% and 78%, respectively, although pancreatic abscesses and those involving yeast were correlated with poor outcomes by this treatment method (10). Khurrum Baig et al. echoed the success of percutaneous drainage in treating abscesses secondary to colorectal surgery, but questioned the applicability of these findings to patients with other than well-defined intra-abdominal abscesses (11). Other considerations include planned relaparotomy and open management. Data is far from optimal, as these critically ill patients cannot ethically be randomized to different treatment groups. However, it would appear at this time that these strategies still result in an unacceptably high mortality of around 42% (12,13). A study by Schein found a particularly high mortality of 55% in the specific subgroup of diffuse postoperative peritonitis treated by planned relaparotomy, with or without open management. Furthermore, Schein went on to state that open management was associated with over twice the mortality of closed: 58% versus 24% (14). Although necessary flaws in study design make it difficult to say whether these approaches offer an advantage over the more traditional ones, it is nevertheless clear that they are far from ideal.

294

Kaiser and Wilson

The hurdles in addressing the challenge of tertiary peritonitis have led to exploration of potential future therapies. Some are in keeping with traditional surgical/ mechanical means: case studies have surfaced detailing the success of laparoscopy, even in the face of diffuse peritonitis and multiple abscesses (15). Other concepts favor a medicine-based approach, rooted in emerging ideas on the disease’s basic pathology. As it is believed that bacteria migrate out of the intestines secondary to mucosal weakening, strategies that strengthen the mucosa, such as early postoperative enteral feeding or selective elimination of endogenous pathogenic bacteria, have each been tried with mixed results. Likewise, it has been argued that the progression from secondary to tertiary peritonitis represents a crippling of the body’s immune system; in support of this belief, granulocyte colony stimulating factor and interferon-c have each produced limited success in small patient groups, and successfully treated individuals all demonstrated some recovery of immune cell functioning. Another idea has been postulated that a relative lack of corticosteroid exists to fulfill the demands of extreme stress, and it has been seen that supplying patients with stress doses of hydrocortisone can dramatically improve the vascular effects of septic shock. Finally, some researchers have investigated the possibility that alleviating the hyper-catabolic state of patients with tertiary peritonitis might decrease mortality. Growth hormone and insulin-like growth factor-1 have both been tried with intermittent positive and negative outcomes (9).

NEW ONSET PERITONITIS Antibiotic-associated Clostridium difficile Diarrhea in the ICU Patient Epidemiology, Pathogenesis, and Risk Factors The anaerobe Clostridium difficile causes twice as many cases of diarrhea as all other bacterial and protozoal causes combined. In hospitalized patients, C. difficile is responsible for 30% of diarrhea cases, and in hospitalized patients receiving antibiotic therapy—as is the case for many postsurgical patients—this number rises to an impressive 50% to 70%. C. difficile–associated diarrhea (CDAD), is theorized to arise in patients colonized by the pathogen when protective normal gut flora is simultaneously suppressed by broad-coverage antibiotic exposure. Although clindamycin, ampicillin, and the third-generation cephalosporins such as ceftazidime, ceftriaxone and cefotaxime are the most commonly associated antimicrobials, the newer, broaderspectrum quinolones such as gatifloxacin and moxifloxacin can also increase risk, and in fact any antibiotic, including, surprisingly, metronidazole and vancomycin, may rarely predispose patients to the disease. Other risk factors for CDAD include age, >60 years, the winter season, antineoplastic agents (especially methotrexate), recent gastrointestinal surgery, enemas, stool softeners, postpyloric enteric tube feedings (e.g., J-tubes), and even use of proton-pump inhibitors in hospitalized patients (16,17). Diagnosis A CDAD diagnosis is reached based on a number of clinical and laboratory findings such as low-grade fever, median leukocytosis of around 16,000 WBCs/mm3, occasional hypoalbuminemia secondary to a protein-losing enteropathy, and, in 5% of patients, even the dramatic presentation of acute abdomen. Sigmoidoscopy, when

Intra-Abdominal Surgical Infections and Their Mimics in the Critical Care Unit

295

performed in equivocal cases, will show whitish or yellowish pseudomembranes overlying the mucosa in 41% of cases, and radiologic studies, although nonspecific, will often show signs of inflammation such as cecal dilatation, air–fluid levels, and mucosal thumbprinting. Even though diagnosis is often confirmed using the enzymelinked immunoassay, it is worth bearing in mind that these tests are only about 85% sensitive. Even polymerase chain reaction (PCR), culture, and the cytotoxicity assay—considered to be the gold-standard in terms of specificity—are likewise imperfect; therefore, a negative test result should not undermine the weight of sound clinical judgment when other likely causes of nosocomial diarrhea have been ruled out (16,17). Treatment and Prevention Therapy for mild cases may consist only of discontinuing the offending antibiotics, or switching to antibiotics less likely to perpetuate CDAD, such as aminoglycosides, macrolides, sulfonamides, or tetracyclines: up to a quarter of cases will resolve following this step alone. For moderate-to-severe cases, metronidazole, either orally or intravenously, is the first line of therapy. In the 20% to 30% of patients who will relapse, a second course of metronidazole is recommended, followed by vancomycin enema for persistent symptomatic infection. Other treatments, such as intravenous immunoglobulin, cholestyramine which binds the bacterial toxin, and probiotics such as Lactobacillus, the yeast Saccharomyces boulardii, and even donor feces or ‘‘stool transplantations’’ to seed the re-growth of normal gut flora, have all been tried with success but as yet are not commonly done. Of course, prevention remains the most effective means of addressing the C. difficile dilemma, and precautions such as contact isolations for known carriers, conscientious hand-washing, gloves, and bleach disinfection of hospital surfaces, endoscopes and other equipment should never be overlooked (16,17). Acalculous Cholecystitis Acalculous cholecystitis, with its difficulty in diagnosis and attendant high mortality, should be a consideration in jaundiced postoperative patients. Although this disease occurs in only about 0.19% of SICU patients, it nevertheless accounts for around 14% of all acute cholecystitis patients, and the mortality ranges from 15% to 41% (18,19). With this in mind, physicians caring for high-risk populations should carefully evaluate the signs and symptoms of this disease, and even a low level of clinical suspicion should prompt more thorough investigation. Risk Factors and Pathophysiology Although the pathogenesis of acalculous cholecystitis has not been entirely elucidated, it is apparent that the critically ill patient is particularly prone. Risk factors include recent trauma, burn injury, or nonbiliary tract operations, atherosclerosis, diabetes, hypertension, chronic renal failure, hemodynamic instability such as congestive heart failure or shock, and use of total parenteral nutrition (TPN) (18–21). One patient has been reported in the literature with acalculous cholecystitis secondary to a diaphragmatic hernia mechanically obstructing the cystic duct (19). Only about 13% have a history indicative of gallbladder disease (21). Given these associations, it is likely that there are multiple triggering factors contributing to a common disease state. An experimental form of the disease is produced by a combination of decreased blood flow to the gallbladder, cystic duct obstruction, and bile concentration (21). It can

296

Kaiser and Wilson

be conjectured that a partially ischemic state, together with the effects of stasis, creates a favorable environment for the growth of enteric bacteria, ultimately leading to inflammation, often with accompanying gangrene, empyema, perforation, and abscess at rates much higher than those seen with calculous cholecystitis (18,20,21). E. coli is the organism most commonly isolated (19). Presentation and Diagnosis In addition to having one or more of the above risk factors, acalculous cholecystitis patients frequently present with the classical signs and symptoms of the calculous form, such as right upper quadrant pain, Murphy’s sign, nausea and vomiting, abdominal distention, decreased bowel sounds, fever, jaundice, and abdominal mass (19,21); although patients with mental status changes often lack pain and other symptoms, absence of any one clue should not exclude such a serious possibility (18,22). Laboratory values suggesting the diagnosis include leukocytosis, hyperamylasemia, and elevated aminotransferases (22). Nevertheless, these findings are nonspecific, and given the likelihood of atypical presentation, the equivocal patient generally warrants radiologic and/or nucleotide (isotope) tests including ultrasound, CT scan, and cholescintigraphy such as hepatobiliary iminodiacetic acid (HIDA) scan. Of these, cholescintigraphy demonstrating an abnormal gallbladder ejection fraction of 250/mm3 may be further supported by positive single organism ascites fluid cultures, this test is only about 60% sensitive even under optimal conditions—bedside aerobic and anaerobic cultures of 10 mL each into blood culture bottles—and moreover requires unacceptable delay as a practical indication of treatment (32). Although recent studies have shown promising results of 100% sensitivity in the diagnosis of SBP using certain urine reagent strips, these findings are not yet supported by sufficient experience to advocate their routine clinical use (37). Secondary peritonitis is bacterial peritonitis secondary to a viscus perforation, surgery, abdominal wall infection, or any other acute inflammation of intra-abdominal organs. In the postsurgical ICU patient, differentiating SBP from secondary peritonitis is particularly challenging, yet nonetheless pivotal in determining appropriate management. Secondary peritonitis often occurs in the wake of obvious causes, but in settings where underlying issues are subtle, a diagnosis of SBP may be mistakenly seized and acted upon. Thus, a diagnosis of secondary peritonitis should generally be considered when patients fail antibiotic therapy for SBP. Characteristics of ascites fluid strongly favoring secondary peritonitis over SBP include isolation of multiple organisms, isolation of anaerobic or fungal organisms, or an ascites glucose level 10 g/L and lactic dehydrogenase concentration greater than that of normal serum. These indicators are all very sensitive but nonspecific for a diagnosis of secondary peritonitis, and their presence must be weighed against the remaining clinical picture before any firm diagnoses are reached (32). Treatment and Prognosis Initial empiric treatment for SBP must cover gram-negative aerobic bacteria from the family of Enterobacteriacae as well as nonenterococcal Streptococcal species, and must adequately penetrate into the peritoneal fluid. Low dose, short course cefotaxime—2 g twice a day for five days—is generally considered the first line therapy, but other cephalosporins such as cefonicid, ceftriaxone, ceftizoxime, and ceftazidime are equally effective, and even oral, lower cost antibiotics such as amoxicillin with clavulanic acid will achieve similar results. For patients with penicillin allergy, oral fluoroquinolones such as ofloxacin are yet another suitable option, except in those with a history of failed quinolone prophylaxis implying probable resistance.

300

Kaiser and Wilson

Follow-up paracentesis is recommended after 48 hours of antibiotic therapy to assess response: a fall >25% in the number of ascites PMN cells is considered a success (32). However, antimicrobials are not the only means of management: because renal impairment secondary to decreased intravascular volume is a major cause of mortality in SBP, further management may be aimed at preventing this fluid shift. The addition of albumin to an antibiotic regimen has been shown to decrease in-hospital mortality almost two-thirds from 28% to 10%. It is considered especially beneficial for patients with already impaired renal function and a creatinine >91 mmol/L, or advanced liver disease as evidenced by serum bilirubin >68 mmol/L (33). Nevertheless, the future outlook for patients with SBP is bleak: of those that survive the initial episode 30% to 50% will survive one year further, and only 25% to 30% will live a second year. Given these odds, patients with a history of SBP should be considered for liver transplantation, as well as long-term antibiotic prophylaxis in the interim (33). Prophylaxis On weighing the cost of antimicrobials and the threat of inducing antibiotic resistance against the gravity of SBP, prophylaxis is indicated only for patients with the highest risk, namely, those with a previous episode of SBP, ongoing gastrointestinal bleeding, or an ascitic fluid protein 60%) of older children and adults have detectable antibodies against C. difficile toxins (6), and serum levels of antitoxin A IgG rise rapidly after colonization, indicating a systemic amnestic response to the toxin (8). Serum IgG and IgA and mucosal IgA all appear to be involved in protection (6). Serum antitoxin A IgG and fecal antitoxin A IgA levels are higher in patients who develop mild CDAD than in those with prolonged, severe diarrhea (8). For patients with severe underlying disease, 88% of those with IgG levels 3.0 enzyme linked immunosorbent assay (ELISA) units developed CDAD, compared to 20% with higher levels; among patients with mild to moderate underlying disease, 43% with low IgG levels developed CDAD, versus none with higher levels. Using multiple logistic regression analysis, the odds of developing CDAD for patients with serum IgG levels 3.0 ELISA units were 48 times greater than for those with IgG levels > 3.0 units, after adjustment for age, sex, and severity of disease (8). Another study found fewer macrophages and IgA-producing cells in patients with CDAD, particularly in those with PMC, compared to controls with non–C. difficile diarrhea (28). Immune response also appears important in relapsing disease; children with recurrent CDAD had lower serum IgG against toxin A than controls (6). There is no evidence that serum antibodies are protective against colonization (8).

Clostridium difficile –associated Diarrhea and Colitis in the Critical Care Unit

309

MICROBIOLOGY C. difficile is a gram-positive, large (2–17 lm), spore-forming anaerobic bacillus. It is closely related to C. sordellii but not to other toxigenic Clostridia, such as C. perfringens, C. botulinum, and C. tetani. Difficult to isolate in the laboratory (hence its name), C. difficile can be grown on highly selective cefoxitin, cycloserine, and fructose agar media (5). C. difficile colonizes the luminal surface of the colon but is generally noninvasive. Outside of the gastrointestinal tract, C. difficile demonstrates low pathogenicity, although it may enhance that of other bacteria in mixed infections (29). C. difficile produces two heat-labile protein exotoxins (toxins A and B), the largest known bacterial toxins (30). These toxins are optimally expressed at body temperature (5). Purified toxins are capable of causing the full spectrum of disease (15). Although most strains produce both toxins, some produce toxin B only but can be equally pathogenic. Nontoxigenic C. difficile strains are not believed to cause human disease, although rare cases of CDAD caused by strains producing neither toxin A nor B have been reported (6). Toxigenic strains are not equally virulent; some strains that clearly possess toxin genes demonstrate low levels of gene transcription, resulting in minimal toxin production (31). Toxin A is a 308-kDa enterotoxin that produces acute inflammation, induces fluid secretion, and causes necrosis of the epithelium in the rabbit ileal loop model (30). Toxin B is a 270-kDa cytotoxin that is more potent than toxin A in tissue culture (6). The toxins appear to act synergistically (15). Both toxins are internalized and inactivate proteins in the Rho subfamily, which regulate the F-actin cytoskeleton. This results in disaggregation of actin, opening the tight junctions between cells, and resulting in characteristic cell rounding and cell death (6,15). Both toxins are also pro-inflammatory, inducing release of cytokines, phospholipase A2, plateletactivating factor (30), tumor necrosis factor–a, and substance P (9). This results in activation of the enteric nervous system, leading to neutrophil chemotaxis and fluid secretion. C. difficile also produces tissue degradation enzymes, such as collagenase and hyaluronidase (1). Some strains also produce an actin-specific binary toxin, which is encoded by the cdtA and cdtB genes and is cytotoxic to Vero cells in culture. Binary toxin has been associated with more severe C. difficile disease, but whether strains possessing binary toxin are truly more pathogenic requires further study (32). One study found that 10.3% (22/214) of toxigenic strains harbored this binary toxin (33), but 65.3% (32/49) of strains in a recent hospital outbreak demonstrated cdtA and cdtB genes (8). AAD can be caused by other enteric pathogens. Although generally recognized as a cause of food poisoning, several studies have found evidence of enterotoxigenic strains of C. perfringens type A in patients with AAD but not in controls. Elderly hospitalized patients seem to be at particular risk (34). As mentioned above, S. aureus was routinely found in cases of PMC prior to the discovery of C. difficile; although it is felt that this largely represents misdiagnosis, as S. aureus may have the potential to cause a similar syndrome. Candida albicans has been found in large quantities (>100,000 organisms/g stool) in patients with AAD who subsequently improved after receiving antifungal therapy, but it remains unclear whether this represents true infection. Salmonella species have also been shown to cause PMC and have been implicated in rare cases of AAD. However, testing for these organisms is not routinely done.

310

Drees and Gorbach

CLINICAL PRESENTATION Most patients exposed to C. difficile, even after antibiotic exposure, become asymptomatically colonized. Colonization rates of 25% to 80% are seen in healthy infants and neonates (1), but clinical illness is rare. For unclear reasons, colonization appears to wane with advancing age, and only 3% of healthy adults are colonized. Colonization increases to 20% to 30% of hospitalized adults (4), but clinical symptoms develop in only one-third of those who become colonized (6). Once colonization is established, the risk of symptomatic CDAD decreases (1). Symptoms can begin within the first day of antibiotic use, or up to six weeks after completion of the antibiotic course (6). Most commonly, symptoms develop within four to nine days (1). Diarrhea is frequently watery or mucoid and may contain blood or be greenish in color. Mild disease is defined as diarrhea without any systemic symptoms such as fever or hemodynamic changes. Moderate disease may result in profuse diarrhea, abdominal distention or pain, fever, tachycardia, and oliguria, but responds readily to volume resuscitation. Pseudomembranes are seen in more advanced disease and are characterized by raised yellow plaques 2 to 10 mm in diameter, frequently with normal intervening mucosa (Fig. 1) (1,28). Histologically, the membranes are composed of inflammatory cells, fibrin, mucin, and cellular debris (28). PMC primarily affects the large bowel, although the small intestine may rarely be involved. Occasionally patients may present without diarrhea but with marked leukocytosis and abdominal pain, due to primary right-sided involvement (6). In the setting of ileus, CDAD should still be considered in absence of diarrhea. Severe or fulminant disease may result in hemodynamic instability requiring pressor support and/or mechanical ventilation (9), occult bleeding, and severe oliguria. Fulminant colitis develops in 1% to 3% of cases and can lead to ileus, toxic

Figure 1 Typical endoscopic findings in Clostridium difficile–associated pseudomembranous colitis with widely disseminated, punctate yellow plaques with normal intervening mucosa.

Clostridium difficile –associated Diarrhea and Colitis in the Critical Care Unit

311

megacolon, intestinal perforation, and death. The first warning sign may be diminishing diarrhea, due to decreased colonic muscle tone. Dallal et al. report that of 44 patients undergoing colectomy for fulminant colitis, five (11%) presented with frank peritonitis, hypotension, or both (9). With appropriate treatment, the overall mortality for PMC is < 1% in most series (35), but mortality as high as 24% has been reported among critically ill patients (36). Among patients requiring surgery, mortality rates after colectomy have ranged from 38% to 80% in small series (9). In one study of patients with fulminant colitis requiring colectomy, the need for preoperative vasopressor support significantly predicted postoperative mortality (9). Other complications of CDAD include hyperpyrexia, transverse volvulus (5), and protein-losing enteropathy, resulting in hypoalbuminemia and anasarca. Reactive arthritis, similar to Reiter’s syndrome caused by other enteric pathogens, may occur one to four weeks after infection (6,37). Extracolonic C. difficile is rare, but case series have described isolation of the organism from pleural fluid, peritoneal fluid, blood, bone, prosthetic joints, wounds (including necrotizing fascitiis), and splenic, vaginal, and perianal abscesses. Generally these infections are polymicrobial, making it difficult to ascertain the contributing pathogenicity of the C. difficile itself. However, pure extracolonic cultures of C. difficile have been described (29,37). Unfortunately, relapsing CDAD is common even with appropriate treatment, occurring in 20% to 25% of infections. Relapse generally occurs 3 to 21 days after completion of anticlostridial therapy and is due to recurrence rather than reinfection (36). However, one study did find that 50% of relapses were due to reinfection with a different strain, rather than recrudescence of the original infecting strain (15). Most will respond to a second course of therapy, but those who have had two or more recurrences have a 65% risk of further relapse (15,36), and 3% to 5% will have more than six relapses. Up to 26 relapses have been described in a single patient (36). Patients with chronic renal insufficiency, high leukocyte counts (15,000 cells/mL), multiple previous episodes of CDAD, community-acquired disease, and who require continued antibiotic therapy have higher risk of relapse (6). Patients who develop PMC may also be at greater risk; in 63 patients with CDAD who underwent flexible sigmoidoscopy, 17 (30.4%) of 56 patients with PMC relapsed, compared to zero of seven patients without PMC (38).

DIFFERENTIAL DIAGNOSTIC CONSIDERATIONS The majority of AAD cases cannot be attributed to any specific microorganism. A number of theories have been advanced to explain the role of antimicrobial drugs in producing diarrhea not associated with C. difficile. By disturbing the normal bowel flora, antibiotics destroy organisms responsible for producing short chain fatty acids, resulting in longer nonabsorbable molecules reaching the colon and an osmotic diarrhea (2). Carbohydrate metabolism is also affected, with similar results. The breakdown of primary bile acids, which are potent colonic secretory agents, may also be affected (4). In addition, certain antibiotics have direct effects on the gastrointestinal system. For example, erythromycin increases the gastric emptying rate, clavulanate stimulates bowel motility, and neomycin causes malabsorption (1). Penicillin has rarely been noted to cause segmental colitis (4). Hospitalized patients are frequently subjected to polypharmacy, and medication lists should be carefully examined for other contributing agents, such as laxatives, antacids, electrolyte supplements

312

Drees and Gorbach

(particularly magnesium), nonsteroidal antiinflammatory drugs (NSAIDS), contrast, products containing lactose or sorbitol, and antiarrhythmic and cholinergic medications. Ulcerative colitis may present similarly to CDAD, and typhlitis should be considered in neutropenic patients (39). Only 2% to 3% of AAD has been proven to be caused by alternate pathogens, such as C. perfringens, S. aureus, and C. albicans (1); routine testing for these organisms is not recommended. Although PMC is much more specific for C. difficile, other potential etiologies include early ischemia, verotoxin-producing organisms such as Escherichia coli 0157:H7, and medications (gold, chlorpropanide, or NSAIDS). Hospitalized patients without history of antibiotic use and without significant amounts of diarrhea or abdominal pain are unlikely to have C. difficile; 94% to 97% of patients meeting these criteria will have negative cytotoxin assays (5). Testing for C. difficile is not recommended in infants under one year, for nondiarrheal stool specimens (except in setting of ileus) or for test of cure (5). Given the wide spectrum of disease caused by C. difficile, there are no pathognomonic findings on history or physical exam. Diarrhea is seen in nearly all cases, but can range from insignificant ‘‘nuisance’’ diarrhea to profuse, cholera-like diarrhea. In more severe disease, abdominal pain, bloating, and tenderness are generally present, and rarely the disease can present as an acute abdomen. One study found that 35% of patients with fulminant colitis caused by C. difficile were diagnosed at autopsy (9), suggesting that a significant number of deaths due to ‘‘sepsis’’ in critically ill patients may be related to C. difficile. Although nonspecific, leukocytosis is common in CDAD and can precede the diarrhea and abdominal pain. Band forms are frequently present. One prospective study of 400 inpatients with white blood cell (WBC) counts 15,000 cells/mL found C. difficile infection in 11% of those with WBC of 15 to 19,900 cells/mL, 15% of those with WBC 20 to 29,000 cells/mL, and 34% of those with WBC 30,000 cells/mL. A retrospective analysis found proven C. difficile in 25% of patients with WBC 30,000 (excluding those with leukemia). Conversely, among 133 outpatients with WBC 15,000, only one (1%) was proven to have C. difficile (40). Other supporting laboratory findings include hypoalbuminemia and fecal leukocytes. Fecal leukocytes have 28% to 40% sensitivity and 92% specificity (5). Fecal lactoferrin assays have been found to have sensitivity of 75% to 90%, but are nonspecific (46%) (5). Radiologic studies are also nonspecific but can support the diagnosis. Plain abdominal films may reveal mucosal edema or paralytic ileus and are useful in ruling out free intraabdominal air and toxic megacolon. Computed tomography (CT) may show diffusely thickened colon with edematous mucosa (6). One study of 39 patients with CDAD who underwent CT found that all were diagnostic when combined with the clinical scenario, showing ascites and colonic wall thickening or massive dilatation. Eleven patients had right-sided colitis, whereas nine had left-sided colitis and 19 had pancolitis (9). Barium enemas are not recommended due to the risk of perforation (39). If endoscopy is performed and pseudomembranes are visualized, the likelihood of C. difficile is high. However, mucosa can be normal or can demonstrate minimal erythema (39). One study of 179 patients with undiagnosed diarrhea who underwent flexible sigmoidoscopy found 63 (35.2%) patients with CDAD, of whom 56 (88.9%) had PMC. Twenty-nine (52%) of these patients had negative cytotoxin assays from stool sampled during sigmoidoscopy; nine of these patients had stool samples available for culture, all of which demonstrated toxigenic C. difficile (38). Another study

Clostridium difficile –associated Diarrhea and Colitis in the Critical Care Unit

313

of 20 patients who underwent flexible sigmoidoscopy for CDAD found a false negative rate of 10%. Of the two patients not diagnosed, one had strictly right-sided disease and the other had poor bowel preparation (9). Endoscopy should be avoided in patients with severe disease with colonic dilatation.

DIAGNOSIS The gold standard for diagnosis of C. difficile infection is the tissue culture cytotoxin assay for toxin B, which can detect as little as 10 pg of toxin (4). The assay reveals characteristic cytopathic effects on cell culture monolayers; preincubation with neutralizing antibodies demonstrates the specificity of the cytotoxicity (Fig. 2). Sensitivity and specificity are high (94–100% and 99%, respectively) (6). However, this test requires tissue culture capability. It has been supplanted largely by cytotoxin enzyme immunoassay (EIA), which requires 100 to 1000 pg of either toxin A or B, and can provide results within hours. The EIA has high specificity (92–98%); sensitivity can be as low as 71%, but can be increased to 90% with testing of three stool samples. More than one assay for diagnosis is required in 5% to 20% of patients (6). False negative assays may be due to toxin instability, degradation of toxin by proteases derived from other bacteria, or the presence of other bacteriaderived toxin-binding components (38). Assays that test for both toxins are preferred, as 1% to 2% of strains produce only toxin B (4). Fatal PMC was reported in a patient with a C. difficile strain that produced nonfunctional toxin A; the

Figure 2 Tissue culture assay for Clostridium difficile. (A) Normal primary human amnion cells. (B) Typical changes (cell rounding) after application C. difficile toxin. (C) Tissue culture after neutralization with Clostridium sordelli antitoxin.

314

Drees and Gorbach

repeated negative cytotoxin A assays resulted in delay of diagnosis and contributed to the patient’s death (41). Stool culture is highly sensitive for C. difficile but has several disadvantages. Because nontoxigenic strains are frequently present, culture must be accompanied by toxin-culture assay and broth culture of isolates to identify toxigenic strains. As a result, diagnosis may be delayed by three to four days, and most laboratories no longer perform C. difficile cultures (4). Latex agglutination assays lack sensitivity and specificity, as they detect the enzyme glutamate dehydrogenase rather than toxin (1). Polymerase chain reaction (PCR) is very sensitive, but requires significant technical expertise. However, a rapid detection method developed in Spain using nested PCR of the toxin B genes has been found to be 96% sensitive and 100% specific and can be performed in several hours (1). This assay is not yet widely available. TREATMENT Whenever possible, the inciting antibiotic should be discontinued or changed. For mild disease, no further treatment may be necessary. Diarrhea may resolve in 25% of patients just with discontinuation of antibiotics (6). Supportive care, such as intravenous fluids and electrolyte replenishment, should be offered if necessary. Antiperistaltic agents, such as narcotics and loperamide, should be avoided as they may promote a dire complication, toxic megacolon (4). Treatment of asymptomatic carriers is ineffective (42), may prolong the carrier state, and is not advised. Asymptomatic carriage usually is transient and resolves spontaneously (39). Indications for treatment include severe diarrhea, persistent diarrhea despite stopping antibiotics, evidence of systemic toxicity, and the need to continue antibiotics (Table 2) (4). Duration of treatment is usually 10 days, although many experts recommend continuing anticlostridial therapy for one week following discontinuation of the inciting antibiotic. Empiric treatment while awaiting cytotoxin Table 2 Treatment of Clostridium difficile–associated Diarrhea General supportive measures Discontinue inciting antibiotic(s) if possible, or change to antibiotic rarely associated with CDAD Correct fluid and electrolyte imbalances Avoid use of antiperistaltic agents Anti-clostridial antibiotics Metronidazole 500 mg orally three times daily for 10 days Vancomycin 125 mg orally four times daily for 10 days (first-line therapy in severely ill, pregnant, or lactating patients) Metronidazole 500 mg intravenously three times daily (in patients with ileus or inability to take oral medications) Relapsing disease Repeat 10-day course of metronidazole or vancomycin Lactobacillus GG: one capsule twice daily for 14 days Saccharomyces boulardii: 500 mg capsules twice daily for four weeks IVIG: 400 mg/kg every three weeks Consider administration of fecal organisms via nasogastric tube or retention enema Abbreviations: CDAD, C. difficile–associated diarrhea; IVIG, intravenous immunoglobulin; GG.

Clostridium difficile –associated Diarrhea and Colitis in the Critical Care Unit

315

results, or if the first assay is negative, is advisable in severely ill patients with suspected C. difficile disease. One study found that patients who died from fulminant colitis were twice as likely as those who survived to have had an initial false-negative toxin (9). Typical response is fairly rapid, with decreased fever within one day and improvement of diarrhea in four to five days. In patients who fail to respond, one should consider lack of compliance, an alternate diagnosis, or the inability of drug to reach the colon, such as with ileus or megacolon (4). Oral metronidazole and oral vancomycin are the two agents most commonly used for CDAD. Currently published guidelines from the Society for Healthcare Epidemiology of America, the American College of Gastroenterology, and the Hospital Infection Control Practices Advisory Committee all recommend metronidazole as the first-line agent (42). Both have similar response rates (90–97%), but metronidazole has much lower cost and vancomycin has the potential to promote colonization with vancomycin-resistant enterococcus. A 10-day course of high-dose oral vancomycin (500 mg four times daily) costs US $7358 compared to $765 for a 10-day course of oral metronidazole (dosed at 500 mg three times daily) (42). Metronidazole is typically dosed orally at 500 mg three times daily or 250 mg four times daily. Metronidazole is well absorbed and difficult to detect in healthy volunteers without diarrhea (43), reaching relatively low fecal concentrations (0.4–24.4 mg/g feces after 400 mg orally or 500 mg intravenously) (35). Parenteral therapy may be required in the setting of severe ileus or toxic megacolon. Intravenous metronidazole is primarily excreted in the upper gastrointestinal tract, and approximately 14% of the dose is excreted in feces (3). However, fecal concentrations of the drug have been shown to exceed the minimum inhibitory concentration (MIC) with parenteral therapy (6). Metronidazole treatment failures occur, although documented drug resistance of C. difficile is rare. However, C. difficile isolates rarely undergo sensitivity testing. One study found no metronidazole resistance even among 10 primary treatment failures (43). Another study found only one isolate with high-level resistance to metronidazole (MIC > 64 mg/mL by agar dilution testing) of 100 tested (35). The mechanism of resistance is not well understood. No antibiotic resistance plasmids have ever been reported in C. difficile (17). The rate of treatment failure may be increasing; in Quebec in 2003, 87 of 110 patients with high leukocyte counts, creatinine levels, or both, were treated initially with metronidazole. Complicated CDAD developed in 34 (39.1%), and 20 (23.0%) died (11). Indications for oral vancomycin therapy include pregnancy, lactation, intolerance of metronidazole, or failure to respond to metronidazole within three to five days of treatment (4). Because of the recent experience with metronidazole-treatment failures, and the low fecal levels achieved with this drug, some experts have moved to treatment of more severely ill patients with vancomycin (39). Parenteral vancomycin is not effective. The typical oral dose is 125 mg four times daily. In the setting of ileus or severe disease, one can increase the dose to 250 to 500 mg four times daily (4). Adjunctive intracolonic vancomycin, administered via retention enema, has been shown to be effective in small, uncontrolled case series of patients with severe or fulminant colitis not responding to standard therapy (3). Vancomycin resistance has not been reported. In the 2003 outbreak in Quebec, initial treatment with vancomycin was associated with a 79% lower risk of complicated CDAD compared with metronidazole, after adjustment for confounding factors (adjusted OR 0.2, 95% CI 0.06–0.8) (11). Alternate agents for the treatment of CDAD include teicoplanin, fusidic acid, and bacitracin (6). Teicoplanin is available in Europe but not in the United States, and has

316

Drees and Gorbach

been touted to be superior to vancomycin in terms of both bacteriologic and symptomatic cure (42). Both fusidic acid and bacitracin have been shown to be less effective than vancomycin (42). Anion exchange resins, such as cholestyramine, bind toxin in the colon but have been associated with treatment failures when used alone. Rifaximin, newly available in the United States for treatment of travelers’ diarrhea, has wide antibacterial activity and poor absorption, leading to high intraluminal concentrations. It was compared in vitro with metronidazole and vancomycin against 93 C. difficile isolates, and demonstrated superior intrinsic activity (44). In small clinical trials it was as effective as vancomycin (42). It is not Food and Drug Administration–approved for the treatment of CDAD, however. A minority of patients (0.39–3.6%) with C. difficile colitis require surgery (45). Surgery is indicated for patients with peritoneal signs, systemic toxicity, toxic megacolon, perforation, multiorgan failure, or progression of symptoms despite appropriate antimicrobial therapy (39,45). Total colectomy with end ileostomy is the procedure of choice. Select patients with disease clearly limited to the ascending colon have been treated successfully with right hemicolectomy, but intraoperative colonoscopy should be performed to rule out left-sided disease (9). Treatment of relapsing disease can become problematic. Most authorities recommend repeating a second 10-day course of metronidazole, with 92% response rates (46). For patients with multiple relapses, a variety of options have been tried: long courses of metronidazole, vancomycin, or both; vancomycin plus rifampin; tapered or pulsed dosing of vancomycin; or vancomycin plus cholestyramine (36). Of note, cholestyramine binds vancomycin as well as C. difficile toxin, so doses should be separated by several hours (46). Recovery of normal fecal flora may take days to weeks after the discontinuation of antibiotics (16). Aside from cost, repeated courses of anticlostridial therapy have the disadvantage of perpetuating this disruption in intestinal flora. To break this cycle, alternate treatments have been attempted, including feces or fecal flora via enema or nasogastric tube, nontoxigenic C. difficile (46), and probiotics. Stool transplantation was shown effective in one small case series. Filtered stool from patients’ family members or healthy volunteers was administered via nasogastric tube to 18 patients with at least two relapses of CDAD. Two patients died from apparently unrelated causes and one patient had one additional relapse, but the remaining patients were cured (overall 94% cure rate). The author notes that the procedure was well tolerated and most patients felt symptomatically better within 24 to 48 hours (36). This study was retrospective and uncontrolled, however. Probiotics are nonpathogenic microorganisms that, when ingested, may benefit the health or physiology of the host (47). Probiotics have been beneficial in the setting of travelers’ diarrhea, rotavirus infection, and AAD (48), but most have only been studied thus far in small, open label, or uncontrolled trials. Organisms include brewers yeast (Saccharomyces cerevisiae), S. boulardii, Lactobacillus GG (LGG), and L. plantarum LP299v (47). S. boulardii was studied in conjunction with metronidazole or vancomycin in patients with multiple recurrences of CDAD, and decreased relapses compared to placebo (35% vs. 65%, p ¼ 0.04) (46). Lactobacilli are a diverse group of lactose-fermenting organisms that have several immune-enhancing effects, including augmentation of phagocyte function and enhancement of humoral and cell-mediated immune responses (48). LGG is a human isolate resembling L. casei subspecies rhamnosus that has inhibitory activity against a wide range of bacteria, including C. difficile. Unlike some other lactobacilli, it can survive digestion and persist in the colon for at least

Clostridium difficile –associated Diarrhea and Colitis in the Critical Care Unit

317

one week (2). Data regarding its efficacy is conflicting, however. One study of 188 children receiving antibiotics and LGG found a significant decrease in antibioticassociated diarrhoea (AAD) (24% in placebo group vs. 7% in LGG group), as well as decreased duration of diarrhea and improved stool consistency (48). Similar findings were reported in a study from Finland (49). However, another prospective, randomized, placebo-controlled trial of 267 adults found no difference in the incidence of diarrhea during a 21-day follow-up period (2). Because the host immune response to C. difficile challenge plays a major role, passive immunotherapy with intravenous IgG (IVIG) has been studied in patients with recurrent or refractory CDAD (8). A dose of 400 mg/kg every three weeks was found to produce a marked increase in serum antitoxin A/B levels, and resolution of diarrhea (46). Five children with relapsing disease and low antitoxin A IgG levels responded favorably to IVIG, and several studies in adults have also shown favorable results (15). PREVENTION Prevention of C. difficile colonization and subsequent infection requires aggressive infection control within hospitals and other institutions (Table 3). Hand washing and use of barrier precautions (gown and gloves) during patient contact requires continual emphasis. Of note, the widely used alcohol-based hand sanitizing solutions do not destroy C. difficile spores; this may be contributing to the recent increases in CDAD described above. When CDAD is suspected, patients should be placed in enteric precautions while laboratory testing is ongoing, until the diagnosis can Table 3 Hospital Infection Control Policies for Clostridium difficile–associated Diarrhea Patient interventions Use enteric precautions (gown and gloves upon entering room) for patients with confirmed or suspected CDAD Place patients with confirmed or suspected CDAD in private rooms and bathrooms Consider diagnosis and order cytotoxin assays promptly Avoid unnecessary use of acid suppression (i.e., proton pump inhibitors) and gastrointestinal motility agents Consider prophylactic use of probiotics in patients at particularly high risk of CDAD Hospital staff interventions Improve hand hygiene compliance Encourage use of soap and water rather than alcohol-based hand sanitizers after contact with C. difficile–infected patients Environmental interventions Clean patient rooms with hypochlorite solutions (1 part bleach to 10 parts water, prepared daily and allowed to air dry) Individually assign thermometers, blood pressure cuffs, and stethoscopes Avoid use of rectal thermometers Adequately disinfect all equipment used by multiple patients Hospital policy interventions Pursue multidisciplinary interventions to improve appropriate antibiotic use May need to restrict high-risk antimicrobials Cohort nursing staff Cohort patients if necessary in outbreak setting Abbreviation: CDAD, Clostridium difficile–associated diarrhea.

318

Drees and Gorbach

be excluded. Patients should be isolated in private rooms (with private bathrooms) or cohorted with other C. difficile–infected patients. Enteric precautions can be removed when diarrhea ceases without test of cure. Patient rooms should be disinfected with freshly prepared hypochlorite solutions (1 part bleach to 10 parts water) and allowed to air dry. Appropriate antibiotic use is vital; one study found 30% of all antimicrobialdays of therapy in a teaching hospital unnecessary (16). Even a few doses of certain antibiotics can cause prolonged disruption of fecal flora, therefore one should strive to avoid improper initiation of antibiotics. During epidemics, hospital-wide restriction of implicated antibiotics has been effective (4). One epidemic linked to a clonal clindamycin-resistant C. difficile isolate responded to restriction of clindamycin; a decrease in the mean number of cases per month from 11.7 to 5.7 was observed within the first six months, with further decline to 3.5 cases per month during the second year of restriction (14). The number of isolates resistant to clindamycin decreased from 91% to 35% (14). The total costs of antibiotics with antianaerobic activity increased, as agents such as imipenem and pipercillin–tazobactam were substituted for clindamycin, but the hospital experienced overall net savings due to decreased number of CDAD cases (14). Even in the absence of an epidemic, restriction of third-generation cephalosporin usage has been associated with decreased rates of C. difficile (19). Vaccination to prevent CDAD has also been studied. Hamster studies have shown protection from lethal ileocecitis with parenteral formalin-inactivated toxins A and B, and full protection from death and diarrhea with a combination of intranasal and intraperitoneal inactivated culture filtrate vaccine plus cholera toxin and Ribi adjuvants. Passive immunization with antitoxin A–neutralizing antibodies protected against lethal disease, whereas full protection against diarrhea required antibodies to both toxins. IgG monoclonal antibodies against the cell-binding domain of toxin A provided complete protection in gnotobiotic mice (15). In humans, an investigational parenteral toxoid vaccine using inactivated toxins A and B was recently tested in healthy volunteers for safety and immunogenicity. The vaccine was well tolerated and all subjected seroconverted (15). Whether the antibody titers elicited are sufficient to protect against disease is unknown. An initial pilot test of three patients with chronic, relapsing CDAD found no recurrent disease during two-month follow-up.

REFERENCES 1. Hurley BW, Nguyen CC. The spectrum of pseudomembranous enterocolitis and antibioticassociated diarrhea. Arch Intern Med 2002; 162:2177–2184. 2. Thomas MR, Litin SC, Osmon DR, et al. Lack of effect of Lactobacillus GG on antibioticassociated diarrhea: randomized, placebo-controlled trial. Mayo Clin Proc 2001; 76:883–889. 3. Apisarnthanarak A, Razavi B, Mundy LM. Adjunctive intracolonic vancomycin for severe Clostridium difficile colitis: case series and review of the literature. Clin Infect Dis 2002; 35:690–696. 4. Bartlett JG. Antibiotic-associated diarrhea. N Engl J Med 2002; 346(5):334–339. 5. Thielman NM, Wilson KH. Antibiotic-associated colitis. In: Mandell GL, Bennett JE, Dolin R, eds. Mandell, Douglas, and Bennett’s Principles and Practice of Infectious Diseases. 6th ed. Philadelphia, Pennsylvania: Elsevier Churchill Livingstone Publications, 2005:1249–1263. 6. Mylonakis E, Ryan ET, Calderwood SB. Clostridium difficile-associated diarrhea: a review. Arch Intern Med 2001; 161:525–533.

Clostridium difficile –associated Diarrhea and Colitis in the Critical Care Unit

319

7. Archibald LK, Banerjee SN, Jarvis WR. Secular trends in hospital-acquired Clostridium difficile disease in the United States, 1987–2001. J Infect Dis 2004; 189:1585–1589. 8. Kyne L, Warny M, Qamar A, et al. Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A. N Engl J Med 2000; 342 (6):390–397. 9. Dallal RM, Harbrecht BG, Boujoukas AJ, et al. Fulminant Clostridium difficile: an under appreciated and increasing cause of death and complications. Ann Surg 2002; 235(3):363–372. 10. Morris AM, Jobe BA, Stoney M, et al. Clostridium difficile colitis: an increasingly aggressive iatrogenic disease? Arch Surg 2002; 137:1096–1100. 11. Pe´pin J, Valiquette L, Alary M-E, et al. Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ 2004; 171(3):466–472. 12. Pindera L. Quebec to report on Clostridium difficile in 2005. CMAJ 2004; 171(7):715. 13. Valiquette L, Low DE, Pe´pin J, et al. Clostridium difficile infection in hospitals: a brewing storm. CMAJ 2004; 171(1):27–29. 14. Climo MW, Israel DS, Wong ES, et al. Hospital-wide restriction of clindamycin: effect on the incidence of Clostridium difficile-associated diarrhea and cost. Ann Intern Med 1998; 128:989–995. 15. Giannasca PJ, Warny M. Active and passive immunization against Clostridium difficile diarrhea and colitis. Vaccine 2004; 22:848–856. 16. Donskey CJ. The role of the intestinal tract as a reservoir and source for transmission of nosocomial pathogens. Clin Infect Dis 2004; 39:219–226. 17. Johnson S, Samore MH, Farrow KA, et al. Epidemics of diarrhea caused by a clindamycinresistant strain of Clostridium difficile in four hospitals. N Engl J Med 1999; 341(22): 1645–1651. 18. Gerding DN. Clindamycin, cephalosporins, fluoroquinolones, and Clostridium difficileassociated diarrhea: this is an antimicrobial resistance problem. Clin Infect Dis 2004; 38:646–648. 19. Thomas C, Stevenson M, Williamson J, et al. Clostridium difficile-associated diarrhea: epidemiological data from Western Australia associated with a modified antibiotic policy. Clin Infect Dis 2002; 35:1457–1462. 20. McCusker ME, Harris AD, Perencevich E, et al. Fluoroquinolone use and Clostridium difficile-associated diarrhea. Emerg Infect Dis 2003; 9(6):730–733. 21. Gaynes R, Rimland D, Killum E, et al. Outbreak of Clostridium difficile infection in a long-term care facility: association with gatifloxacin use. Clin Infect Dis 2004; 38:640–645. 22. Safdar N, Maki DG. The commonality of risk factors for nosocomial colonization and infection with antimicrobial-resistant Staphylococcus aureus, Enterococcus, gram-negative bacilli, Clostridium difficile, and Candida. Ann Intern Med 2002; 136:834–844. 23. Emoto M, Kawarabayashi T, Hachisuga T, et al. Clostridium difficile colitis associated with cisplatin-based chemotherapy in ovarian cancer patients. Gynecol Oncol 1996; 61:369–372. 24. Husain A, Aptaker L, Spriggs DR, et al. Gastrointestinal toxicity and Clostridium difficile diarrhea in patients treated with paclitaxel-containing chemotherapy regimens. Gynecol Oncol 1998; 71:104–107. 25. Kent KC, Rubin MS, Wroblewski L, et al. The impact of Clostridium difficile on a surgical service: a prospective study of 374 patients. Ann Surg 1998; 227(2):296–301. 26. Zimmaro Bliss D, Johnson S, Savik K, et al. Acquisition of Clostridium difficile and Clostridium difficile-associated diarrhea in hospitalized patients receiving tube feeding. Ann Intern Med 1998; 129:1012–1019. 27. Dial S, Alrasadi K, Manoukian C, et al. Risk of Clostridium difficile diarrhea among hospital inpatients prescribed proton pump inhibitors: cohort and case-control studies. CMAJ 2004; 171(1):33–38. 28. Johal SS, Lambert CP, Hammond J, et al. Colonic IgA producing cells and macrophages are reduced in recurrent and non-recurrent Clostridium difficile associated diarrhea. J Clin Pathol 2004; 57:973–979.

320

Drees and Gorbach

29. Wolf LE, Gorbach SL, Granowitz EV. Extraintestinal Clostridium difficile: 10 years’ experience at a tertiary-care hospital. Mayo Clin Proc 1998; 73:943–947. 30. Guerrant RL, Steiner TS, Lima AAM, et al. How intestinal bacteria cause disease. J Infect Dis 1999; 179(suppl 1):S331–S337. 31. Philips C. Serum antibody responses to Clostridium difficile toxin A: predictive and protective? Gut 2001; 49:167–168. 32. McEllistrem MC, Carman RJ, Gerding DN, et al. A hospital outbreak of Clostridium difficile disease associated with isolates carrying binary toxin genes. Clin Infect Dis 2005; 40:465–472. 33. Gonc¸alves C, Decre´ D, Barbut F, et al. Prevalence and characterization of a binary toxin (actin-specific ADP-ribosyltransferase) from Clostridium difficile. J Clin Microbiol 2004; 42(5):1933–1939. 34. Modi N, Wilcox MH. Evidence for antibiotic induced Clostridium perfringens diarrhea. J Clin Pathol 2001; 54:748–751. 35. Wong SS-Y, Woo PC-Y, Luk W-K, et al. Susceptibility testing of Clostridium difficile against metronidazole and vancomycin by disk diffusion and etest. Diagn Microbiol Infect Dis 1999; 34:1–6. 36. Aas J, Gessert CE, Bakken JS. Recurrent Clostridium difficile colitis: case series involving 18 patients treated with donor stool administered via a nasogastric tube. Clin Infect Dis 2003; 36:580–585. 37. Jacobs A, Barnard K, Fishel R, et al. Extracolonic manifestations of Clostridium difficile infections: presentation of 2 cases and review of the literature. Medicine 2001; 80(2): 88–101. 38. Johal SS, Hammond J, Solomon K, et al. Clostridium difficile associated diarrhea in hospitalized patients: onset in the community and hospital and role of flexible sigmoidoscopy. Gut 2004; 53:673–677. 39. Yassin SF, Young-Fadok TM, Zein NN, et al. Clostridium difficile-associated diarrhea and colitis. Mayo Clin Proc 2001; 76:725–730. 40. Wanahita A, Goldsmith EA, Musher DM. Conditions associated with leukocytosis in a tertiary care hospital, with particular attention to the role of infection caused by Clostridium difficile. Clin Infect Dis 2002; 34:1585–1592. 41. Johnson S, Kent SA, O’Leary KJ, et al. Fatal pseudomembranous colitis associated with a variant Clostridium difficile strain not detected by toxin A immunoassay. Ann Intern Med 2001; 135:434–438. 42. Bricker E, Garg R, Nelson R, et al. Antibiotic treatment for Clostridium difficileassociated diarrhea in adults. Cochrane Database Systematic Rev 2004 (most recent update; Accessed Feb 2005). 43. Sanchez JL, Gerding DN, Olson MM, et al. Metronidazole susceptibility in Clostridium difficile isolates recovered from cases of C. difficile-associated disease treatment failures and successes. Anaerobe 1999; 5:201–204. 44. Marchese A, Salerno A, Pesce A, et al. In vitro activity of rifaximin, metronidazole, and vancomycin against Clostridium difficile and the rate of selection of spontaneously resistant mutants against representative anaerobic and aerobic bacteria, including ammoniaproducing species. Chemotherapy 2000; 46:253–266. 45. Synnott K, Mealy K, Merry C, et al. Timing of surgery for fulminating pseudomembranous colitis. Br J Surg 1998; 85:229–231. 46. Kyne L, Kelly C. Recurrent Clostridium difficile diarrhoea. Gut 2001; 49:152–153. 47. Marteau PR, de Vrese M, Cellier CJ, et al. Protection from gastrointestinal diseases with the use of probiotics. Am J Clin Nutr 2001; 73(suppl):430S–436S. 48. Vanderhoof JA. Probiotics: future directions. Am J Clin Nutr 2001; 73(suppl):1152S– 1155S. 49. Arvola T, Laiho K, Torkkeli S, et al. Prophylactic Lactobacillus GG reduces antibioticassociated diarrhoea in children with respiratory infections: a randomized study. Pediatrics 1999; 104(5):e64.

16 Severe Skin and Soft Tissue Infections in the Critical Care Unit Mamta Sharma and Louis D. Saravolatz Division of Infectious Disease, Department of Medicine, St. John Hospital and Medical Center, and Wayne State University School of Medicine, Detroit, Michigan, U.S.A.

INTRODUCTION Skin and soft tissue infections are common and vary widely in severity from minor pyodermas to severe necrotizing infections. Most of these infections are superficial and treated with regimens of local care and antimicrobial therapy. However, others like necrotizing infections are life threatening and require a combined medical and surgical intervention. Prompt recognition and treatment is paramount in limiting the morbidity and mortality associated with these infections, and thus a thorough understanding of the various etiologies and presentation is essential in the critical care setting. It is also important to discriminate between infectious and noninfectious causes of skin and soft tissue inflammation. A detailed history and examination are necessary to narrow the possible etiologies of infection. In many instances surface cultures are unreliable and misleading because surface colonizing organisms can be mistaken for pathogens. In instances in which the diagnosis is in doubt, aspiration, biopsy, or surgical exploration of the skin can be considered. Here we review causes of severe skin and soft tissue infection, highlighting the clinical presentation, diagnosis, and approach to management in the critical care setting.

MICROBIAL FLORA Physiological factors that control the bacterial skin flora include humidity, water content, skin lipids, temperature, and rate of desquamation. The pH of the skin is usually around 5.6. Besides containing secretory Immunoglobulin (IgA), sweat also possesses sufficient salt to create a high osmotic pressure, which may be responsible for inhibiting many microbial species. In spite of these barriers to colonization, the skin provides an excellent venue of various microenvironments. Differences in cutaneous microflora may relate to variability in skin surface temperature and moisture content as well as the presence of different concentrations of skin surface lipids that 321

322

Sharma and Saravolatz

may be inhibitory to various microorganisms. Colonization with organisms sensitive to desiccation, such as gram-negative bacilli, is not favored. The predominant bacterial flora of the skin is the various species of coagulase-negative staphylococci (S. epidermidis, S. capitis, S. warneri, S. hominis, S. haemolyticus, S. lugdunensis, and S. auricularis), Corynebacterium species (diphtheroids), and Propionibacterium species. Humans are a natural reservoir for Staphylococcus aureus (S. aureus), and asymptomatic colonization is far more common than infection. Colonization of the nasopharynx, perineum, or skin, particularly if the cutaneous barrier has been disrupted or damaged, may occur shortly after birth and may recur anytime thereafter (1–4). The anterior nares are reservoirs for S. aureus. Approximately 20% of individuals always carry one type of strain and are called persistent carriers. A large proportion of the population, approximately 60%, harbors S. aureus intermittently, and the strains change with varying frequency. Such persons are called intermittent carriers. Finally, approximately 20% almost never carry S. aureus and are called noncarriers (5–7). Carriage rates are higher than in the general population for injection drug users, persons with insulin-dependent diabetes, patients with dermatologic conditions, patients with long-term indwelling intravascular catheters, and those with human immunodeficiency virus infection. High nasal carriage rates are found in patients with S. aureus skin infections as demonstrated from nasal cultures taken at the time the S. aureus infection was present (5). Micrococcus spp., Peptostreptococcus, Streptococcus viridans, and Enterococcus spp. can also be isolated. Acinetobacter spp. are found on the skin of about 25% of the population in the axillae, toe webs, groin, and antecubital fossa. Proteus, Pseudomonas, Enterobacter, and Klebsiella are rarely found. Antibiotics disturb the balance within commensal flora and leave the surface vulnerable to colonization by exogenous gram-negative bacilli and fungi. The principal fungal flora is lipophilic yeasts of the genus Malassezia, and nonlipophilic yeasts such as Candida spp. are also inhabitants of the skin (1,2,4). Primary skin infections occur in otherwise normal skin and are usually caused by group A streptococci or S. aureus. Secondary infections complicate chronic skin conditions (e.g., eczema or atopic dermatitis). A deficiency in the expression of antimicrobial peptides may account for the susceptibility of patients with atopic dermatitis to skin infection with S. aureus (8). These underlying disorders act as a portal of entry for virulent bacteria. Other factors predisposing to skin infections include vascular insufficiency, disrupted venous or lymphatic drainage, sensory neuropathies, diabetes mellitus, previous cellulitis, foreign bodies, accidental or surgical trauma, burns, poor hygiene, obesity, and immunodeficiencies.

CLASSIFICATION OF SKIN AND SOFT TISSUE INFECTIONS Infections of the skin and soft tissue can be divided based on the depth of penetration and the ability of the organism to produce necrosis. Infection of the outermost layer of skin, the epidermis, is termed impetigo. Extension into the superficial dermis with involvement of lymphatics is typical of erysipelas, whereas cellulitis is an extension into the subcutaneous tissue. In necrotizing fasciitis (NF) there is involvement of fascia, whereas in myonecrosis there is involvement of muscle. A clinically useful distinction with important management implications subdivides soft tissue infections into non-necrotizing and necrotizing processes (9). In some systemic infections, cutaneous manifestations are non-infectious complications of the illness as in purpura fulminans.

Severe Skin and Soft Tissue Infections in the Critical Care Unit

323

Impetigo Impetigo is the most common, contagious, superficial skin infection produced by S. aureus or streptococcus. There are two clinical presentations: bullous impetigo and nonbullous impetigo, and both begin as a vesicle (10). Bullous impetigo, like staphylococcal scalded skin syndrome (SSSS) and the staphylococcal scarlatiniform syndrome, represents a form of cutaneous response to the two extracellular exfoliative toxins produced by S. aureus of phage group II (usually type 71). The group A streptococci responsible for impetigo belong to different M serotypes (2,49,52,55,57, 59–61) from those of strains that produce pharyngitis (1,2,4,6,25) (11,12). Crusted impetigo is usually associated with a mixed flora of both S. aureus and streptococci. S. aureus is known to be the primary pathogen in both bullous and nonbullous impetigo. They are common in exposed areas, such as hands, feet, and legs, and are often associated with traumatic events, such as minor skin injury or insect bite. Predisposing factors include warm ambient temperature, humidity, poor hygiene, and crowded conditions. Systemic complications are very uncommon.

Staphylococcal Scalded Skin Syndrome SSSS, first described in 1956, is a generic term applied to a group of exfoliative dermopathies caused by an exfoliative (or epidermolytic) exotoxin, produced by various strains of S. aureus, mainly of phage group II (usually type 71) (13–15). It primarily affects neonates and young children, although adults with underlying diseases are also susceptible. Two variants of the toxin, the exfoliative toxin A and B, have been described. These exotoxins induce pathological changes in the epidermis that closely resemble a scald caused by boiling water, hence the name SSSS (16–18). Histologically, these toxins cause intraepidermal cleavage through the granular layer without damage or alteration of the keratinocytes, bullae formation, and slippage of the upper epidermal layer with the application of gentle pressure (a positive Nikolsky’s sign). S. aureus enterotoxin (A through D) and toxic shock syndrome toxin 1 (TSST-1) are frequently associated with staphylococcal scarlet fever. The clinical response to these exotoxins is varied. Thus, the manifestations of SSSS include several primarily age-dependent presentations: (i) a generalized exfoliative syndrome seen in newborns (Ritter’s disease or Pemphigus neonatorum) and children, but can rarely develop in adults; (ii) bullous impetigo, a localized pustulosis in children; and (iii) staphylococcal scarlet fever, a form of SSSS that does not progress beyond the initial stage of a generalized erythematous eruption. SSSS occurs abruptly or few days after a recognized staphylococcal infection with fever, skin tenderness, and scarlatiniform rash. The lesions begin as a vesicle that gradually enlarges into flaccid bullae that rupture, leaving a tender, moist surface that eventually heals. Localized infection occurs, usually in the nasopharynx, umbilicus, or urinary tract. Large flaccid clear bullae form over two to three days, and result in separation of sheets of skin. Exfoliation exposes large areas of bright red skin surface (19,20). Fluid and electrolyte loss can lead to hypovolemia and sepsis syndrome. In adults the mortality rate approaches 60% (21). With appropriate therapy the lesions heal within two weeks. Toxic epidermal necrolysis (TEN) typically occurs as a drug reaction. The lesions are similar to SSSS; however, there is more extensive destruction of the epidermis and the stratum corneum layer, recovery is prolonged, and scarring is more frequent. TEN is often fatal and should be treated like a widespread burn. Most cases of SSSS are diagnosed on clinical grounds and

324

Sharma and Saravolatz

are easily treated with antibiotics, which rapidly eliminate the staphylococci producing the toxin. Laboratory investigations are required only if the clinical findings are equivocal or when outbreaks occur. Because the condition is the result of exotoxins, which may be produced by staphylococci at a distant site, the blister fluid in generalized SSSS tends to be sterile, whereas the fluid in localized bullous impetigo will contain S. aureus. Staphylococci producing enterotoxions ET can usually be cultured from the nares, conjunctiva, or nasopharynx. Biopsy of the blister is one of the most definitive diagnostic tests in SSSS. One study revealed a positive blister biopsy result with intraepidermal cleavage in all 30 adults with SSSS (20). Blood cultures are usually negative because the organisms are frequently noninvasive, particularly in children. In one study only 3% of children had a positive blood culture, in contrast to 20 (62.5%) of 32 adults (17,20,22–24). Treatment: Severe forms require more aggressive treatment with intravenous antistaphylococcal antibiotics and extra care of denuded skin to prevent secondary infection and fluid losses, and to maintain body temperature, especially in neonates. In methicillin sensitive strains, a penicillinase-resistant penicillin nafcillin or oxacillin (2 g IV every four to six hours) is the drug of choice. Cefazolin (1–2 g IV every eight hours) is an alternative treatment that can also be used in patients with histories of delayed type penicillin allergy. In methicillin-resistant strains (MRSA) vancomycin (1 g or 15 mg/kg IV every 12 hours), sulfamethaxole/trimethoprim (1600/320 IV every 12 hours), linezolid (600 mg IV or orally every 12 hours), and other agents such as daptomycin (4 mg/kg/day IV) for skin and soft tissue infections (6 mg/kg/day IV for severe infections) and quinupristin-dalfopristin (7.5 mg/kg IV every eight hours) are treatment options (25,26). Linezolid, daptomycin, and quinupristin-dalfopristin can be used for vancomycin intermediate S. aureus and vancomycin-resistant S. aureus (VRSA) strains (25). Oritavancin, dalbavancin, tigecyline, and telavancin are newer agents under development for treatment of resistant strains (27). Toxic Shock Syndrome Toxic-shock syndrome (TSS) is a rapid-onset illness causing fever, hypotension, rash, multiple organ system dysfunction, and desquamation. Infection with S. aureus produces classical TSS, whereas Streptococcus pyogenes causes a modified form of TSS known as either streptococcal TSS (STSS), or toxic-shock–like syndrome (TSLS). TSLS displays many of the typical TSS symptoms with the addition of severe soft tissue necrosis (28). Diagnosis of TSLS caused by streptococci is based on a constellation of clinical and laboratory signs as proposed by the Centers for Disease Control and Prevention (Table 1) (29,30). There are two clinical forms of TSS: menstrual TSS and nonmenstrual TSS. Menstrual TSS starts within three days of the beginning or end of menses and is primarily associated with the use of high absorbency tampons. Clinical signs include high fever, capillary leak syndrome with hypotension and hypoalbunemia, generalized nonpitting edema, and a morbilliform rash, followed by desquamation after a few days. TSST-1 and staphylococcal enterotoxins are the paradigm of a large family of pyrogenic exotoxins called superantigens (SAg). For nonmenstrual TSS, the offending pathogen can virtually colonize any site in the body (31–34). Recurrent menstrual TSS is a well-described phenomenon (35,36). Two conditions are required for recurrence of TSS: persistent colonization with a toxigenic strain of S. aureus and persistent absence of neutralizing antibody. Recurrent TSS develops exclusively among patients who fail to develop a humoral immune response to the implicated

Severe Skin and Soft Tissue Infections in the Critical Care Unit

325

Table 1 Streptococcal Toxic-Shock Syndrome: Clinical Case Definition An illness with the following clinical manifestations occurring within the first 48 hrs of hospitalization or, for a nosocomial case, within the first 48 hrs of illness: Hypotension defined by a systolic blood pressure less than or equal to 90 mmHg for adults or less than the fifth percentile by age for children aged less than 16 yrs. Multiorgan involvement characterized by two or more of the following: Renal impairment: creatinine greater than or equal to 2 mg/dL (greater than or equal to 177 mmol/L) for adults or greater than or equal to twice the upper limit normal for age. In patients with preexisting renal disease, a greater than twofold elevation over the baseline level Coagulopathy: platelets less than or equal to 100,000/mm3 (less than or equal to 100  106/L) or disseminated intravascular coagulation, defined by prolonged clotting times, low fibrinogen level, and the presence of fibrin degradation products Liver involvement: alanine aminotransferase, aspartate aminotransferase, or total bilirubin levels greater than or equal to twice the upper limit of normal for the patient’s age. In patients with preexisting liver disease, a greater than twofold increase over the baseline level Acute respiratory distress syndrome defined by acute onset of diffuse pulmonary infiltrates and hypoxemia in the absence of cardiac failure or by evidence of diffuse capillary leak manifested by acute onset of generalized edema, or pleural or peritoneal effusions with hypoalbuminemia A generalized erythematous macular rash that may desquamate. Soft tissue necrosis, including necrotizing fasciitis or myositis, or gangrene Laboratory criteria for diagnosis Isolation of group A Streptococcus. Case classification Probable: a case that meets the clinical case definition in the absence of another identified etiology for the illness and with isolation of group A Streptococcus from a nonsterile site Confirmed: a case that meets the clinical case definition and with isolation of group A Streptococcus from a normally sterile site (e.g., blood or cerebrospinal fluid, or less commonly, joint, pleural, or pericardial fluid)

staphylococcal toxin (37). Diagnosis of TSS is based on a constellation of clinical and laboratory signs as proposed by the Centers for Disease Control and Prevention (Table 2) (28). In the late 1980s, a disease similar in appearance to TSS, yet caused by invasive streptococci, was recognized and referred to as ‘‘toxic strep,’’ ‘‘streptococcal TSLS,’’ or STSS. This condition was found to share many clinical features with TSS. M types 1, 3, 12, and 28 have been the most common isolates from patients with shock and multiorgan failure (38,39). In the majority of cases toxin-producing group A streptococci have been isolated, with Streptococcal pyrogenic exotoxin-A (Spe-A) production being most closely linked with invasive disease. However, group A streptococci producing Streptococcal pyrogenic exotoxin-B (Spe-B), Streptococcal pyrogenic exotoxin-C (Spe-C), streptococcal SAg, and mitogenic factor, as well as nongroup A streptococci, have been found to be causative in individual cases of STSS as well. Similar to classic TSS, the clinical signs of STSS are postulated to be mediated by massive cytokine release (primarily TNF-alpha, IL-1b, and IL-6) as a result of toxin/SAg activity, in addition, streptolysin O, produced by 100% of streptococcal strains associated with STSS, has also been shown to cause TNF-a and IL-1 b production, and has been demonstrated to act synergistically with Spe-A (40–45). Very young, elderly, diabetic, or immunocompromised persons are more susceptible to the acquisition of invasive streptococcal

326

Sharma and Saravolatz

Table 2 Toxic Shock Syndrome: Clinical Case Definition An illness with the following clinical manifestations: Fever: temperature greater than or equal to 102.0 F (greater than or equal to 38.9 C) Rash: diffuse macular erythroderma Desquamation: 1–2 wks after onset of illness, particularly on the palms and soles Hypotension: systolic blood pressure less than or equal to 90 mmHg for adults or less than fifth percentile by age for children aged less than 16 yrs; orthostatic drop in diastolic blood pressure greater than or equal to 15 mmHg from lying to sitting, orthostatic syncope, or orthostatic dizziness Multisystem involvement (three or more of the following): Gastrointestinal: vomiting or diarrhea at onset of illness Muscular: severe myalgia or creatine phosphokinase level at least twice the upper limit of normal Mucous membrane: vaginal, oropharyngeal, or conjunctival hyperemia Renal: blood urea nitrogen or creatinine at least twice the upper limit of normal for laboratory or urinary sediment with pyuria (greater than or equal to five leukocytes per high-power field) in the absence of urinary tract infection Hepatic: total bilirubin, alanine aminotransferase enzyme, or aspartate aminotransferase enzyme levels at least twice the upper limit of normal for laboratory Hematologic: platelets less than 100,000/mm3 Central nervous system: disorientation or alterations in consciousness without focal neurologic signs when fever and hypotension are absent Laboratory criteria Negative results on the following tests, if obtained: Blood, throat, or cerebrospinal fluid cultures (blood culture may be positive for Staphylococcus aureus) Rise in titer to Rocky Mountain spotted fever, leptospirosis, or measles Case classification Probable: a case which meets the laboratory criteria and in which four of the five clinical findings described above are present Confirmed: a case which meets the laboratory criteria and in which all five of the clinical findings described above are present, including desquamation, unless the patient dies before desquamation occurs

infection such as STSS. However, the majority of cases of STSS have occurred in young, otherwise healthy persons between 20 and 50 years of age. An absence of protective immunity is postulated as a potential risk factor in this population. STSS has also been well described as a complication of wounds, varicella, and influenza A. A controversial association of invasive group A streptococcal infections such as STSS with prior nonsteroidal anti-inflammatory drug (NSAID) use has been suggested (46). The link has been proposed to be depression of the cellular immune response by NSAIDs. Clinically, STSS shares many features with TSS. Fever, hypotension, myalgias, liver abnormalities, diarrhea, emesis, renal dysfunction, and hematologic abnormalities may be present in TSS caused by either staphylococci or streptococci. Diffuse macular erythroderma likewise is frequently present in disease caused by both bacteria and is often accompanied by mucous membrane findings, such as conjunctival injection and delayed desquamation of palms and soles. Nonetheless, certain important differences exist between STSS and TSS. The skin is often the portal of entry in STSS, with soft tissue infections developing in 80% of patients (38). The initial presentation of STSS is often localized pain in an extremity, which rapidly progresses over 48 to 72 hours to manifest both local

Severe Skin and Soft Tissue Infections in the Critical Care Unit

327

and systemic signs of STSS. Cutaneous signs may include localized edema and erythema, a bullous and hemorrhagic cellulitis, NF or myositis, and gangrene. Soft tissue involvement of this nature is distinctly uncommon in staphylococcal TSS. Blood cultures are positive in 60% of patients with STSS (38), as compared with less than 3% in TSS. Mortality in STSS is between 30% and 80%, whereas in staphylococcal TSS ranges from 3% to 5% (47,48). Treatment: Group A streptococcus is susceptible to penicillin and other b-lactam antibiotics in vitro; however clinical treatment failure occurs when penicillin is used alone in severe group A streptococcus infections (49). This may be attributed to the large inoculum size, the so-called Eagle effect (50,51). These large inoculum reach the stationary growth phase very quickly. Penicillin and other b-lactam antibiotics are ineffective in the stationary growth phase because of reduced expression of penicillin-binding proteins in this phase. Moreover, toxin production is not inhibited by b-lactam antibiotics during the stationary growth phase. The greater efficacy of clindamycin is multifactorial: it inhibits protein synthesis, and its efficacy is unaffected by inoculum size or the stage of bacterial growth. Clindamycin also suppresses synthesis of penicillin-binding proteins, and has a longer post antibiotic effect than b-lactam antibiotics. Lastly clindamycin causes suppression of lipopolysaccharides (LPS)-induced monocyte synthesis of TNF (51–54). Prompt antimicrobial therapy with high dose penicillin and clindamycin should be instituted. Aggressive fluid resuscitation is needed because of intractable hypotension and diffuse capillary leak. Human polyspecific intravenous immunoglobulin (IVIG) has been suggested as a potential adjunctive therapy for invasive group A streptococcus diseases mainly because of its ability to neutralize a wide variety of SAg and to facilitate opsonization of streptococci. An observational cohort study of IVIG in patients with STSS reported decreased mortality rates in patients treated with IVIG compared to controls (67% vs. 34%) (55). A double blind placebo trial was prematurely terminated because of slow recruitment. Analyses of primary end point revealed a reduced mortality in IVIG treated group as compared with placebo (10% vs. 36%), although statistical significance was not reached. A significant increase in plasma neutralizing activity against SAgs expressed by autologous isolates was noted in the IVIG group after treatment (56). If IVIG is to be used, it should be given early and more than one dose should be used, because batches of IVIG have variable neutralizing activity (57). In addition, prompt surgical exploration and debridement of deep seated streptococcal infection should be performed (see discussion of NF). For management of TSS anti-staphylococcal agents are selected with consideration of susceptibility testing. Supportive care includes aggressive intravenous fluid resuscitation and vasopressors as needed. The suspected focus of infection requires specific attention. Specifically, management includes the removal of any vaginal device in menstrual cases and the removal of packed dressings in conjunction with drainage and debridement in cases associated with postsurgical wounds. Furuncles and Carbuncles Furuncle is a deep inflammatory nodule that develops from predisposing folliculitis. A carbuncle is a more extensive process that extends into the subcutaneous fat in areas covered by thick, inelastic skin. Multiple abscesses separated by connective tissue septa develop and drain to the surface along the hair follicle. S. aureus is the most common etiological agent. Infections occur in areas that contain hair follicles such as neck, face, axillae, and buttocks, and sites predisposed to friction and perspiration.

328

Sharma and Saravolatz

Predisposing factors include obesity, defects in neutrophil dysfunction, and diabetes mellitus. Bacteremia can occur and result in osteomyelitis, endocarditis, or other metastatic foci. Erysipelas Erysipelas is a distinctive superficial cellulitis of the skin with prominent lymphatic involvement. In typical erysipelas, the area of inflammation is raised above the surrounding skin; there is a distinct demarcation between involved and normal skin, and the affected area has a classic orange peal (peau d’orange) appearance. The induration and sharp margin distinguish it from the deeper tissue infection of cellulitis, in which the margins are not raised and merge smoothly with uninvolved areas of the skin (Fig. 1). Systemic signs of chills and fever are common. Flaccid bullae filled with clear fluid may develop on the second or third day. Occasionally the infection spreads more deeply and causes cellulitis, abscess, and NF. Desquamation may occur in 5 to 10 days, and scarring is very uncommon. Erysipelas is almost always caused by group A streptococcus, though streptococci of groups G, C, and B, and rarely S. aureus can also be responsible. Formerly the face was commonly involved but now up to 85% of cases occur on the legs and feet largely due to lymphatic venous disruptions (11,58). Erysipelas can spread rapidly if not treated promptly. Blood cultures are positive in only about 5% of cases (58).

Figure 1 Facial erysipelas involving the right cheek. Sharp demarcation between the erythema and right cheek is evident.

Severe Skin and Soft Tissue Infections in the Critical Care Unit

329

Treatment: There has never been a documented report of group A streptococci resistant to penicillin, and thus penicillin remains the drug of choice, penicillin G 200,000 U every six hours. Other alternative agents include first generation cephalosporins or clindamycin. Agents such as erythromycin and the other macrolides are limited by their rates of resistance, and the fluoroquinolones are generally less active than the b-lactam antibiotics against b-hemolytic streptococci.

Cellulitis Cellulitis is an acute, spreading pyogenic inflammation of the dermis and subcutaneous tissue (59,60). S. aureus and group A b-hemolytic streptococcus species are the common organisms (Fig. 2). Cellulitis commonly begins as erythema, edema, and pain and lacks demarcation. It often occurs in the setting of local skin trauma from skin bite, abrasions, surgical wounds, contusions, or other cutaneous lacerations. Edema also predisposes patients to cellulitis. Specific pathogens are suggested when infections follow exposure to seawater (Vibrio vulnificus) (61,62), fresh water (Aeromonas hydrophila) (63), or aquacultured fish (Streptococcus iniae) (64). Lymphedema may persist after recovery from cellulitis or erysipelas and predisposes patients to recurrences. In addition, spread to adjacent structures may result in osteomyelitis. Cellulitis infrequently occurs as a result of bacteremia. Uncommonly, pneumococcal cellulitis occurs on the face or limbs in patients with diabetes mellitus, alcohol abuse, systemic lupus erythematosus, nephritic syndrome, or a hematological cancer (65). Meningococcal cellulitis occurs rarely, although it may affect both children and adults (66). Bacteremic cellulitis due to V. vulnificus with hemorrhagic bullae may follow the ingestion of raw oysters by patients with cirrhosis, hemochromatosis, or thalassemia. Cellulitis caused by gram-negative organisms usually occurs through a cutaneous source in an immunocompromised patient but can also develop through bacteremia. Cryptococcus neoformans, Fusarium, Proteus, and Pseudomonas spp have been associated with bloodstream infections. Immunosuppressed patients are particularly susceptible to the progression of cellulitis from regional to systemic infections.

Figure 2 Cellulitis of the left thigh in an alcoholic patient blood; cultures grew group B streptococcus.

330

Sharma and Saravolatz

The distinctive features including the anatomical location and the patient’s medical and exposure history should guide appropriate antibiotic therapy. Periorbital cellulitis involves the eyelid and periocular tissue and should be distinguished from orbital cellulitis because of complications of the latter: decreased ocular motility, decreased visual acuity, and cavernous-sinus thrombosis. Diagnostic studies: Diagnosis is generally based on clinical and morphologic features of the lesion. Culture of a needle aspirate is not generally indicated because of a low yield. Among 284 patients, a likely pathogen was identified in 29%. Of 86 isolates, only three represented mixed culture. Gram-positive organisms (mainly S. aureus, group A or B streptococci, and Enterococcus faecalis) accounted for 79% of cases; the remainder were caused by gram-negative bacilli (Enterobacteriaceae, H. influenzae, P. multocida, Pseudomonas aeruginosa, and Acinetobacter spp.) (59). Bacteremia is uncommon in cellulitis with only 2% to 4% yielding a pathogen (59). Blood cultures appear to be positive more frequently with cellulitis superimposed on lymphedema. Radiography and computed tomography are of value when clinical setting suggests a subjective osteomyelitis or there is clinical evidence to suggest adjacent infections such as pyomyositis or deep abscesses. When it is difficult to differentiate cellulitis from NF, a magnetic resonance imaging (MRI) may be helpful, although surgical exploration for a definite diagnosis should not be delayed when the latter condition is suspected. Treatment: Because most cases are caused by streptococci and S. aureus, b-lactam antibiotics with activity against penicillinase-producing S. aureus are the usual drugs of choice. Specific treatment for bacterial causes is warranted after an unusual exposure (human or animal bite or exposure to fresh or salt water), in patients with certain underlying conditions (neutropenia, splenectomy, or immunocompromise), or in the presence of bullae and is described in Table 3. Erysipeloid The localized cutaneous infection caused by Erysipelothrix rhusiopathiae presents as a subacute cellulitis (termed ‘‘erysipeloid’’). It is usually due to contact with fish, shellfish, or infected animals. Contact with this pathogen may occur in recreational settings, domestic exposures, or after lacerations among abattoirs or chefs (67). Lesions are slightly raised and violaceous. Other organisms that cause skin and skin structure infections following exposure to water and aquatic animals include Aeromonas, Plesiomonas, Pseudallescheria boydii, and V. vulnificus. Mycobacterium marinum can also cause skin infection, but this infection is characterized by a more indolent course. For erysipelothrix bacteremia or endocarditis, penicillin G (12 million–20 million units IV daily) is the drug of choice; alternative antimicrobials include ciprofloxacin, cefotaxime, or imipenem-cilastatin. Bites Each year, several million Americans are bitten by animals, resulting in approximately 10,000 hospitalizations. Ninety percent of the bites are from dogs and cats, and 3% to 18% of dog bites and 28% to 80% of cat bites become infected, with occasional sequelae of meningitis, endocarditis, septic arthritis, and septic shock. Animal or human bites can cause cellulitis due to skin flora of the recipient of the bite or the oral flora of the biter. Severe infections develop after bites as a result of hematogenous spread or undetected penetration of deeper structures. In a prospective,

Severe Skin and Soft Tissue Infections in the Critical Care Unit

331

Table 3 Antimicrobiol Therapy and Pathogens Associated with Specific Risk Factors Risk factor Dog and cat bites

Human bites

Salt water

Fresh water or use of leeches Butcher, fish handler or veterinarian

Intravenous drug users

Pathogen Pastcurella multocida and other Pastcurella spp. Staphylococcus aureus, Capnocytophaga, streptococcus Neisseria canis, H. felix, Capnocytophaga canimorsus, anaerobes Eikenella corrodens, anaerobes, S. aureus, Streptococcus viridans Vibrio vulnificus

Recommended therapy

Alternative therapy

Ampicillin/sulbactam Ciprofloxacin 500 mg 1.5–3 g IV q.i.d PO or 400 mg IV b.i.d þ clindamycin 600–900 mg IV t.i.d

Ampicillin/sulbactam Ciprofloxacin 500 mg 1.5–3 g IV q.i.d PO or 400 mg IV b.i.d þ clindamycin 600–900 mg IV t.i.d

Cefotaxime 1–2 g IV Doxycycline 200 mg b.i.d or IV followed by 100–200 mg IV b.i.d ciprofloxacin 500 mg PO or 400 mg IV b.i.d Aeromonas species Ciprofloxacin 400 mg Imipenam/cilastatin IV b.i.d 500 mg IV q.i.d Erysipelothrix Penicillin G 12–20 Ciprofloxacin or rhusiopathiae million units IV cefotaxime or every 4 hrs daily imipenam/cilastatin 500 mg IV q.i.d Linezolid 600 mg PO MRSA, Pseudomonas Vancomycin 1 g IV or IV aeruginosa b.i.d þ ceftazidime b.i.d þ tobramycin 1–2 g IV t.i.d or 5.0/kg/daya or cefepime 1–2 g IV b.i.d ciprofloxacin

Note: Dose to be adjusted for azotemia except for ceftriaxone, doxycycline, clindamycin, and linezolid. a Based on once a day dose of 5.0 mg/kg, however can be given as 1.7 mg/kg IV t.i.d. Abbreviation: MRSA, methicillin-resistant Staphylococcus aureus.

multicenter study of infected dog and cat bites Pasteurella spp. was the most common isolate from both dog bites (50%) and cat bites (75%). Pasteurella canis was the most common isolate of dog bites and P. multocida subspecies the most common isolate of cat bites. Other common aerobes include Streptococci, Staphylococci, Moraxella, and Neisseria. Common anaerobes include Fusobacterium, Bacteroides, Porphyromonas, and Prevotella. Capnocytophaga canimorsus is an invasive organism usually occurring in immunosuppressed patients after a dog bite (68,69). Human bites are usually associated with mixed aerobic and anaerobic organisms including S. viridans and other streptococci, S. aureus, Eikenella corrodens, Fusobacterium, and Prevotella. Clenched fist injuries may lead to infection, tendon tear, joint disruption, or fracture (70). For treatment refer to Table 3.

332

Sharma and Saravolatz

Ecthyma Gangrenosum Ecthyma gangrenosum is the classic skin lesion associated with P. aeruginosa infection in granulocytopenic patients (71–73), and has been reported in 2% to 28% of patients with pseudomonas bacteremia. Rarely this lesion may be caused by other organisms, including S. aureus, Aeromonas, Serratia, Klebsiella, Escherichia coli, Capnocytophaga, Aspergillus, and Candida. Neutropenic patients with overwhelming septicemia develop a patchy dermal and subcutaneous necrosis. The characteristic skin lesion starts with erythematous macular eruptions that become bullous with central ulceration and necrosis. These are usually multiple, occurring in different stages of development, which may concentrate on the extremities or the head and neck. Diagnosis of the etiologic agent may occur with biopsy of the lesion being cultured or isolated from blood cultures. Treatment is primarily by administration of intravenous antimicrobial therapy and by debridement of multiple lesions, which may lessen the bacterial burden. Chancriform Lesions: Anthrax A bioterrorism-associated anthrax outbreak occurred suddenly in the United States in 2001. Out of the 22 cases 11 had the cutaneous form (74). After incubation of one to eight days, a painless, sometime pruritic, papule develops on an exposed area. The lesion enlarges and becomes surrounded by a wide zone of brawny, erythematous, gelatinous, and nonpitting edema. As the lesion evolves it becomes hemorrhagic, necrotic, and covered by an eschar. Frequently lymphadenopathy is present. If untreated bacteremic dissemination can occur. Incision and debridement should be avoided because it increases the likelihood of bacteremia (75). A skin biopsy after the initiation of antibiotics can be done to confirm the diagnosis by culture, polymerase chain reaction, or immunohistochemical testing. With the concern that strains may have been modified to be resistant to penicillin, treatment with ciprofloxacin or doxycycline has been recommended (76). Purpura Fulminans Purpura fulminans is an acute illness most commonly associated with meningococcemia but also seen with pneumococcal or staphylococcal disease (77,78). It is typically characterized by disseminated intravascular coagulation (DIC) and purpuric skin lesions. There are four primary features of this syndrome: large purpuric skin lesions, fever, hypotension, and DIC. However, five cases associated with S. aureus strains have been reported from the Minneapolis-St. Paul, Minnesota, metropolitan area. These strains produced high levels of TSST-1, staphylococcal enterotoxin serotype B, or staphylococcal enterotoxin serotype C. Only two of the five patients survived (79). Staphylococcal purpura fulminans may be a newly emerging illness associated with SAg production. There are no specific guidelines for the therapeutic management of this serious manifestation other than assuring that anti-staphylococcal agents are selected with consideration of susceptibility testing. Necrotizing Cellulitis Infectious gangrene is a cellulitis that rapidly progresses, with extensive necrosis of subcutaneous tissues and the overlying skin. Pathologic changes are those of necrosis

Severe Skin and Soft Tissue Infections in the Critical Care Unit

333

and hemorrhage of the skin and subcutaneous tissue. In most instances, necrotizing cellulitis has developed secondary to introduction of the infecting organism at the site of infection. Streptococcal gangrene is a rare form caused by group A streptococci that occurs at the site of trauma but may occur in the absence of an obvious portal of entry. Cases may follow infection at an abdominal operative wound, around an ileostomy or colostomy, at the exit of a fistulous tract, or in proximity to chronic ulceration. The organisms responsible include Clostridium, Bacteroides, and Peptostreptococcus. The diagnosis is suggested when gas is present or when necrosis develops rapidly in an area of cellulitis. Gram stain and culture of skin drainage, aspirate fluid, or surgical specimens should reveal the pathogenic organisms (80–82). Treatment consists of immediate surgical exploration beyond the involved gangrenous and undermined tissue. Areas of cutaneous necrosis are excised. Repeat exploration is commonly performed within 24 hours. Antibiotic therapy should be guided by Gram-stain results or empirically consist of high dose intravenous penicillin G (3–4 million units every four hours) or ampicillin (2 g every four hours), with the addition of clindamycin. Necrotizing Fasciitis NF is a rapidly spreading infection that involves the fascia and subcutaneous tissue with relative sparing of underlying muscle. The mortality of this disease remains alarmingly high ranging from 6% to 76% (83). Delayed diagnosis and delayed debridement have been shown to increase mortality. Type 1 NF is polymicrobial with at least one anaerobic species isolated in combination with one or more facultative anaerobic species such as nontypable streptococci and enterobacteriaceae (Fig. 3). Type 1 NF is common in postoperative infections and includes Fournier’s gangrene. Type 2 NF is typically monomicrobial, most often caused by group A streptococcus (84), Clostridium perfringens. Other organisms that have rarely been implicated in monobacterial infections include Serratia marcescens, Flavobacterium odoratum, Ochrobactrum anthropi, V. vulnificus, and group G streptococcus and S. aureus (85). NF presents either as an acute and life-threatening condition usually caused by group A streptococcus or clostridium spp., or as a subacute process, usually caused by mixed aerobic and anaerobic organisms. The primary site is the superficial fascia. Bacteria proliferate within the superficial fascia and elaborate enzymes and toxins. The precise mechanism of spread has not been fully elucidated but has been attributed to the expression of hyaluronidase, which degrades the fascia. The key pathological process resulting from this uncontrolled proliferation of bacteria is angiothrombotic microbial invasion and liquefactive necrosis of the superficial fascia. As this process progresses, occlusion of perforating nutrient vessels to the skin causes progressive skin ischemia. This event is responsible for the cutaneous manifestations. As the condition evolves, ischemic necrosis of the skin ensues with gangrene of subcutaneous fat, dermis and epidermis, manifesting progressively as bullae formation, ulceration, and skin necrosis. In early stages (stage 1 NF) the disease is indistinguishable from severe soft tissue infection such as cellulitis and erysipelas and presents with only pain tenderness and warm skin. Margins of the skin are poorly defined with tenderness extending beyond the apparent area of involvement. Blister or bulla formation is an important diagnostic clue. It signals the onset of skin ischemia (stage 2 NF). The late stage (stage 3 NF) signals the onset of tissue necrosis and is characterized by hemorrhagic bullae, skin anesthesia, and gangrene. Systemic manifestation such as fever, hypotension, and multiorgan failure can occur

334

Sharma and Saravolatz

Figure 3 Necrotizing fasciitis of left arm and shoulder in a patient with intravenous drug user (IVDU) who injected in the left arm. Patient underwent disarticulation. One set of blood cultures grew Gemella morbillorum and second set grew Streptococcus constellatus. Operative cultures obtained from left arm grew Klabsiella oxytoca, Peptostreptococcus micros, and P. prevoti.

(86–89). The effects are classically caused by SAg produced by group A streptococcus. NSAIDs are postulated to potentiate tissue damage by decreasing granulocyte adhesion and phagocytosis and increasing cytokine production. NF is a clinical diagnosis with corroborative operative findings that include the presence of grayish necrotic fascia, a lack of resistance of normally adherent superficial fascia, a lack of bleeding of the fascia during dissection, and the presence of foul smelling ‘‘dishwater pus.’’ Features reported to be indicative of NF on the computed tomography scan include deep fascial thickening, enhancement, and fluid and gas in the soft tissue planes. Negative deep fascial involvement on MRI effectively excludes NF. Fine needle aspiration, frozen section of tissue biopsy, fascial biopsy, and skin biopsy for histopathology are all useful in diagnosis of NF. The lack of bleeding may be seen or murky dishwater pus exudates may ooze from the incision site. Pathognomonic for NF is a positive ‘‘finger’’ test. The finger test can be used to delineate the extent of infection into the adjacent normal appearing skin. A 2-cm incision down to the deep fascia is made under local anesthesia. Probing of the level of the superficial fascia is then performed. The lack of bleeding, foul smelling dishwater pus, and minimal tissue resistance to finger dissection constitute a positive finger test, which is diagnostic of NF (90). If a diagnosis of NF is made, emergent surgical debridement and/or fasciotomy should be considered. Debridement beyond the visible margin of infection is necessary. Repeated debridements may be required and should continue until the subcutaneous tissue can no longer be separated from the deep fascia. Fasciotomy may be performed at the time of debridement. If infection progresses despite serial debridements and antibiotics, amputation may be life saving. A combination of broad-spectrum antibiotics, such as a penicillin, and an aminoglycoside or a third generation cephalosporin, and clindamycin or metronidazole can be started depending on the clinical presentation. Once the Gram-stain, culture and sensitivity results are obtained, the antibiotic regimen can be altered based on these findings. The use of IVIG as an adjunctive treatment for patients with STSS has been used on the basis of retrospective studies and one small prospective

Severe Skin and Soft Tissue Infections in the Critical Care Unit

335

randomized trial, but conclusive evidence supporting its use remains limited. IVIG contains many antibodies, which neutralize the exotoxins/SAgs secreted by the streptococcus and are involved in the pathogenesis of STSS. Because STSS and NF are mediated by the streptococcal toxins and inflict their tissue destruction via some of the same cytokines, it was postulated that IVIG would be as effective a treatment in NF as it was in STSS. This has yet to be conclusively demonstrated in a clinical trial. For treatment refer to Table 4. Fournier’s Gangrene It originates as a necrotic black area on the scrotum. It is a fulminant, rapidly progressive subcutaneous infection of the scrotum and penis, which spreads along fascial planes and may extend to the abdominal wall. More than 60% of the patients have diabetes mellitus. Fournier’s gangrene occurs commonly without a predisposing event or after uncomplicated hemorrhoidectomy. Less commonly this can occur after urological manipulation or as a late complication of deep anorectal suppuration. Fournier’s gangrene is characterized by necrosis of the skin and soft tissues of the scrotum and or perineum that is associated with a fulminant, painful, and severely toxic infection (91,93) (91,92). The infection is usually polymicrobial. Successful treatment is again based on early recognition and vigorous surgical debridement. Empiric antibiotic treatment is appropriate until culture results are available. Infection is often polymicrobial. The therapeutic benefit of hyperbaric oxygen treatment remains controversial in this as well as other forms of NF. Clostridium Myonecrosis (Gas Gangrene) C. perfringens type A is the most common organism. Although initial growth of the organism occurs within the devitalized anaerobic milieu, acute invasion and destruction of healthy, living tissue rapidly ensues. Historically, clostridial myonecrosis was a disease associated with battle injuries, but 60% of cases now occur after trauma. It is a destructive infectious process of muscle associated with infections of the skin and soft tissue. It is often associated with local crepitus and systemic signs of toxemia, which are formed by anaerobic, gas forming bacilli of the clostridium species. The infection most often occurs after abdominal operations on the gastrointestinal tract; however, penetrating trauma and frostbite can expose muscle, fascia, and subcutaneous tissue to these organisms. Common to all these conditions is an environment containing tissue necrosis, low oxygen tension, and sufficient nutrients (amino acids and calcium) to allow germination of clostridial spores. The systemic manifestations of gas gangrene are related to the elaboration of potent extracellular protein toxins, especially the a-toxin, a phospholipase C, and h-toxin, a thiol-activated cytolysin (94–97). Clostridia are gram-positive, spore forming, obligate anaerobes that are widely found in soil contaminated with animal excreta. They may be isolated from the human gastrointestinal tract and from the skin in the perineal area. C. perfringens is the most common isolate (present in 80% of cases) and is among the fastestgrowing clostridial species, with a generation time, under ideal conditions, of eight minutes. This organism produces collagenases and proteases that cause widespread tissue destruction, as well as a-toxin, which have a role in the high mortality associated with myonecrosis. The a-toxin causes extensive capillary destruction and hemolysis, leading to necrosis of the muscle and overlying fascia, skin, and subcutaneous tissues. Patients complain of sudden onset of pain at the site of trauma or

336

Sharma and Saravolatz

Table 4 Antimicrobiol Therapy and Microbiology Associated with Diabetic Foot Infection and Necrotizing Fasciitis Clinical syndrome Diabetic foot infection

Type 1 NF

Type 2 NF

Pathogen

Recommended therapy

Alternative therapy

Staphylococcus aureus, Streptococcus Enterobacteriaceae, Pseudomonas aeruginosa anaerobes (Bacteroides, Peptostreptococcus) Anarobes (Bacteroides, Peotostreptococcus) Escherichia coli, Enterobacter, Klebseilla, Proteus

Ampicillin/sulbactam 1.5–3 g IV q.i.d or piperacillin/ tazobactam 3.375 g IV q.i.d or ceftriaxone 1–2 g IV q.d or ciprofloxacin þ metronidazole 500 mg IV or PO t.i.da Ampicillin 1–2 g IV every 4–6 hr p þ gentamicin þ metronidazole 0.5 g t.i.d or q.i.d or clindamycin 900 mg t.i.d

Imipenam/cilastatin 500 mg IV q.i.d or clindamycin þ ciprofloxacin 500–750 mg PO or 400 mg IV b.i.d or cefepime 2 g IV b.i.d

Group A streptococcus

Penicillin 2–3 mu IV every 3–4 hr þ clindamycin 900 mg IV t.i.d  IVIG

Imipenam/cilastatin 500 mg q.i.d or ampicillin/ sulbactam þ gentamicin 5.0 mgb or pipercillin/ tazobactam 3.375 g IV q.i.d Cefazolin 1–2 g IV t.i.d or vancomycin þ clindamycin 900 mg IV t.i.d

Note: Dose adjusted for azotemia except for ceftriaxone, doxycycline, clindamycin and linezolid. a When methicillin-resistant Staphylococcus aureus suspected use vancomycin, linezolid or other active agents. b Based on once a day dose of 5.0 mg/kg/day, however can be given as 1.7 mg/kg IV t.i.d. Abbreviations: NF, necrotizing fasciitis; IVIG, intravenous immunoglobulins.

surgical wounds, which rapidly increases in severity. The skin becomes edematous and tense. Hemorrhagic bullae are common, as is a thin watery, foul smelling discharge. Examination of the wound discharge reveals abundant large, box-car shaped gram-positive rods with a paucity of surrounding leukocytes. The usual incubation period between injury and the onset of clostridial myonecrosis is two to three days but may be as short as six hours. A definitive diagnosis is based on the appearance of the muscle on direct visualization by surgical exposure. Initially, the muscle is pale, edematous, and unresponsive to stimulation. As the disease process continues, the muscle becomes frankly gangrenous, black, and extremely friable. This occurs with septicemia and shock. Nearly 15% of patients have positive blood cultures. Serum creatinine phosphokinase levels are always elevated with muscle involvement. The mortality rate associated with gas gangrene approaches 60%. Among the signs that predict a poor outcome are leukopenia, thrombocytopenia, hemolysis, and severe renal failure. Myoglobinuria is common and can contribute significantly to worsening of renal function. Frank hemorrhage may be present and is a harbinger of DIC. Successful treatment of this life-threatening infection depends on early recognition and debridement of all devitalized and infected tissues. When extremities are involved, amputation is frequently indicated. The role of hyperbaric oxygen therapy has not been established (100% oxygen at 3 atm), but it may have a role early in the treatment of seriously ill patients (98,99). The mainstay of treatment is surgical debridement,

Severe Skin and Soft Tissue Infections in the Critical Care Unit

337

and this should not be delayed. A less life-threatening form of this disease is known as clostridium cellulitis. In this process, the bacterial tissue invasion is primarily superficial to the fascial layer, without muscle involvement. Prompt recognition and treatment as described earlier can reduce the associated morbidity and mortality. High dose penicillin G is the drug of choice. Protein synthesis inhibitors such as combining clindamycin with penicillin have had considerably better efficacy than penicillin alone. Clostridium septicum bacteremia is associated with underlying colon cancer or neutropenic enterocolitis (100). Clostridium sordelli has been reported to cause rapidly progressive myonecrosis with fulminant shock syndrome, particularly in obstetric patients. Black tar heroin use has resulted in outbreak of Clostridium botulinism, C. tetani, and C. sordelli in intravenous drug users. Nonclostridial Myonecrosis Nonclostridial myonecrosis encompasses at least five relatively distinct entities that differ from gas gangrene in their pathogenesis, clinical features and bacteriology: Streptococcal myositis  NF type 2 (see earlier discussion under NF), synergistic nonclostridial anaerobic myonecrosis  NF type 1 (see earlier discussion under NF), anaerobic streptococcal myonecrosis, A. hydrophila myonecrosis, and infected vascular gangrene. Anaerobic streptococcal myonecrosis clinically resembles subacute clostridial gas gangrene. The involved muscles are discolored; in contrast to gas gangrene, early cutaneous erythema is prominent. If not treated, the infection progresses to gangrene and shock. The infection is usually mixed, anaerobic streptococci with group A streptococcus or S. aureus. Treatment involves the use of high dose penicillin and antistaphylococcal agent, if indicated, and surgical debridement. Rapidly progressive myonecrosis resembling clostridial gangrene but caused by A. hydrophila may occur after injuries sustained in freshwater, or in conjunction with medicinal leech therapy. Cellulitis often develops within 12 to 24 hours, accompanied by excruciating pain, marked edema, and bullae. Bacteremia is often documented. Treatment requires prompt antimicrobial therapy and wide surgical debridement. Infected vascular gangrene is a focal, usually indolent, and primarily ischemic process in the small muscles of a distal lower extremity already gangrenous from arterial insufficiency. Diabetic patients are prone to develop this complication, which usually does not extend beyond the area of vascular gangrene to involve viable muscle. Proteus spp., Bacteroides spp., and anaerobic streptococci are among the bacteria found in such lesions (101,102).

DIABETIC FOOT INFECTION This term defines any inframalleolar infection in a person with diabetes mellitus. These include paronychia, cellulitis, myositis, abscesses, NF, septic arthritis, tendonitis, and osteomyelitis. The most common lesion requiring hospitalization is the infected diabetic foot ulcer (Fig. 4). Neuropathy plays a central role, with disturbances of sensory, motor, and autonomic functions leading to ulcerations due to trauma or excessive pressure on a deformed foot. This wound may progress to become actively infected, and by contiguous extension the infection can involve deeper tissues. This sequence can be rapid, especially in an ischemic limb. Various immunologic disturbances, especially involving the polymorphonuclear leukocytes, may affect some

338

Sharma and Saravolatz

Figure 4 (A) Limb threatening left diabetic foot ulcer. (B) Rapid progression to gas gangrene. Patient underwent below knee amputation. Operative cultures grew group G streptococcus, methicillin-resistant Staphylococcus aureus, Streptococcus viridans, Enterococcus spp., and Bacteroides fragilis.

diabetic patients. S. aureus and the b-hemolytic streptococci (groups A, C, G, and especially group B) are the most commonly isolated pathogens. Chronic wounds develop a more complex colonizing flora, including enterococci, enterobacteriaceae, obligate anaerobes, P. aeruginosa, and other nonfermentative gram-negative rods. Hospitalization, surgical procedures, and prolonged antibiotics predispose patients to colonization and infection with MRSA or vancomycin-resistant enterococcus. Community-acquired cases of MRSA are becoming more common. Finally, the two reported cases of VRSA involved a diabetic patient with a foot infection (103–105). Therapy: Initial therapy is empirical and should be based on severity of infection and available microbiological data, such as recent culture results or current smear findings from adequately obtained specimens. The microbiology can be

Severe Skin and Soft Tissue Infections in the Critical Care Unit

339

identified by culture only if specimens are collected and processed properly. Deep tissue specimens, obtained aseptically at surgery, contain the true pathogens more often than do samples obtained from superficial lesions. A curettage, or tissue scraping with a scalpel, from the base of a debrided ulcer provides more accurate results. An antibiotic regimen should always include an agent active against staphylococci and streptococci. Previously treated or severe cases may need extended coverage that also includes commonly isolated gram-negative bacilli and Enterococcus spp. Necrotic, gangrenous, or foul smelling wounds usually require antianaerobic therapy. For moderate to severe infection ampicillin/sulbactam or piperacillin/tazobactam can be used. For life-threatening infections imipenam/cilastin may be a consideration. A high prevalence of MRSA may require use of vancomycin, or other appropriate agents against these organisms. The duration of treatment for life-threatening infection may be two weeks or longer. Many infections require surgical procedures that range from drainage and excision of infected and necrotic tissues to revascularization or amputation. For treatment refer to Table 4.

SKIN AND SOFT TISSUE INFECTIONS IN INJECTION DRUG USERS The mechanism by which infection is established probably relates to tissue trauma, direct effects of drugs, tissue ischemia, and inoculation of bacteria. As a result of repeated injections into a single site, skin and surrounding tissue are damaged, develop local ischemia and necrosis, and become susceptible to infection. Opiates suppress T-cell functions and also inhibit phagocytosis, chemotaxis, and killing by neutrophils and macrophages. Infection ranges from cellulitis to skin and soft tissue abscesses, and occasionally fasciitis and pyomyositis. The most common sites of involvement correspond to injection sites: the upper and lower extremities, the groin and antecubital fossa, with the microbiology being monomicrobial or polymicrobial involving S. aureus, S. viridans, S. pyogenes, Streptococcus anginosus group, E. corrodens, anaerobic organisms like clostridium spp. and Prevotella, and gram-negative enteric organisms including E. coli, Klebsiella, Proteus mirabilis, Pseudomonas, and Enterobacter (106–108). Black tar heroin use has resulted in outbreaks of C. botulinism, C. tetani, and C. sordelli in intravenous drug users. For treatment refer to Table 3.

PYOMYOSITIS Pyomyositis is an infection of the skeletal muscle predominantly caused by S. aureus and Streptococcus spp. (109,110). Other rare organisms include enterobacteriaceae and anaerobic bacteria. Case reports of Aspergillus fumigatus, C. neoformans, Mycobacterium tuberculosis, and M. avium-intracellulare have been reported (111,112). It was originally recognized in patients who acquired the disease in the tropics. Predisposing conditions include diabetes mellitus, cirrhosis, immunosuppressive illness, and HIV, and has been reported in intravenous drug abusers. Presumed pathogenesis involves a prior bacteremia, commonly transient. Bacterial infection of the muscle usually occurs after a penetrating wound, vascular insufficiency, or a contiguous spread. Common muscle involvement includes deltoid, psoas, biceps, gastrocnemius, gluteal, and quadriceps, though any muscle group can be involved. Patients will typically present with fever, pain, tenderness, and swel-

340

Sharma and Saravolatz

Figure 5 Right leg abscess, cultures grew methicillin-resistant Staphylococcus aureus (community acquired).

ling of the involved muscle. Bacteremia is present in 5% to 35% of cases. The diagnosis is best established by computed tomography scan or MRI. Treatment consists of drainage (percutaneous or open-incision). Initial antibiotics should consist of intravenous administration of a b-lactamase–resistant penicillin. Initial vancomycin therapy should be considered if MRSA is suspected. Early modification of initial antimicrobial therapy is based on Gram-stain and culture results.

COMMUNITY-ACQUIRED METHICILLIN-RESISTANT S. AUREUS Community-acquired MRSA has become increasingly endemic in many parts of the world (113,114). The most common clinical syndrome has been skin and soft tissue infections (Fig. 5). S. aureus has been a very uncommon cause of NF, but in a recent study 14 patients were identified as community-acquired MRSA with clinical and intraoperative findings of NF, necrotizing myositis, or both (115). Unfortunately, there are no obvious epidemiologic clues to this etiologic agent, and one sees patients with no prior antimicrobial therapy developing this infection. The organism appears somewhat unique in its characteristics by possessing the staphylococcal cassette chromosome Mec IV gene for methicillin resistance and the Panton Valentine Leukocidin genes encoding for a toxin presumably responsible for necrosis in soft tissue sites as well as the lungs. This organism has prompted many clinicians to add vancomycin, sulfamethoxazole/trimethoprim, linezolid, daptomycin, or other agents effective against MRSA in the empiric treatment of skin and soft tissue infections.

SUMMARY A wide variety of skin and soft tissue infections can occur in the critical care settings. The rise in immunocompromised patients such as those with AIDS, transplant

Severe Skin and Soft Tissue Infections in the Critical Care Unit

341

recipients, and those receiving chemotherapy or prolonged corticosteroid therapy have led to diverse etiologies, clinical manifestations, and severity. S. aureus remains the most common pathogen causing infections from minor skin lesions to severe lifethreatening illness such as purpura fulminans. However a variety of other pathogens may be identified and need to be considered with certain epidemiologic clues. Community-acquired MRSA has become increasingly prevalent in many parts of the world. The most common clinical syndrome has been skin and soft tissue infection, but in a recent study 14 patients were identified as community-acquired MRSA with clinical and intraoperative findings of NF, necrotizing myositis, or both. Important considerations when evaluating patients include underlying medical conditions; exposure history; and presenting signs, symptoms, and radiographic patterns. The key to treating serious skin and soft tissue infections successfully is prompt recognition, followed by appropriate antibiotic and surgical intervention as needed to decrease the morbidity and mortality.

REFERENCES 1. Roth RR, James WD. Microbiology of the skin: resident flora, ecology, infection. J Am Acad Dermatol 1989; 20:367–390. 2. Roth RR, James WD. Microbiol ecology of the skin. Annu Rev Microbiol 1988; 42:441–464. 3. Greene JN. The microbiology of colonization, including techniques for assessing and measuring colonization. Infect Control Hosp Epidemiol 1996; 17:114–118. 4. Granto PA. Pathogenesis and indigenous microorganisms of humans. In: Murray PR, Banon EJ, Jorgersas JH, Pyaller MA, Yolken RH, eds. Manual of Clinical Microbiology. 8th ed. Washington, DC: ASM Press, 2003; 1:44–54. 5. Kluytmans J, Belkun AV, Verbrugh H. Nasal carriage of Staphylococcus aureus: epidemiology, underlying mechanisms, and associated risks. Clin Microbiol Rev 1997: 505–520. 6. Eiff CV, Becker K, Machka K, et al. Nasal carriage as a source of Staphylococcus aureus bacteremia. N Engl J Med 2001; 344:11–16. 7. Williams REO. Healthy carriage of Staphylococcus aureus: its prevalence and importance. Bacteriol Rev 1963; 27:56–71. 8. Ong PY, Ohtake T, Brandt C, et al. Endogenous antimicrobial peptides and skin infections in atopic dermatitis. N Engl J Med 2002; 347:1151–1160. 9. Lewis RT. Soft tissue infections. World J Surg 1998; 22:146–151. 10. Dajani AS, Ferrieri P, Wannamaker LW. Natural history of impetigo. II. Etiologic agents and bacterial interactions. J Clin Invest 1972; 51:2863. 11. Bisno AL, Stevens DL. Streptococcal infections of skin and soft tissues. N Engl J Med 1996; 334:240–245. 12. Scales J, Fleischer A, Krowchuk D. Bullous impetigo. Arch Pediatr Adolesc Med 1997; 151:1168–1169. 13. Melish ME, Glasgow LA. The staphylococcal scalded-skin syndrome. N Engl J Med 1970; 282:1114–1119. 14. Dajani AS. The scalded skin syndrome: relation to phage group II staphylococci. J Infect Dis 1972; 125:548–551. 15. Kondo IS, Sakuri S, Sarai Y, et al. Two serotypes of exfoliatin and their distribution in Staphylococcal strain isolated from patients with scalded skin syndrome. J Clin Microbiol 1975; 5:397–400. 16. Yamasaki O, Yamguchi T, Sugai M, et al. Clinical manifestations of staphylococcal scalded-skin syndrome depend on serotypes of exfoliative toxins. J Clin Microbiol 2005; 43:1890–1893.

342

Sharma and Saravolatz

17. Ladhani S, Joannou CL, Lochrie DP, et al. Clinical, microbial, and biochemical aspects of the exfoliative toxins causing staphylococcal scalded-skin syndrome. Clin Microbiol Rev 1999; 12:224–242. 18. Lina G, Gillet Y, Vandenesch F, et al. Toxin involvement in staphylococcal scalded skin syndrome. Clin Infect Dis 1997; 25:1369–1373. 19. Elias, PM, Fritsch P, Epstein EH Jr. Staphylococcal scalded skin syndrome. Arch Dermatol 1997; 113:207–219. 20. Ladhani S, Evans, RW. Staphylococcal scalded skin syndrome. Arch Dis Child 1998; 78: 85–88. 21. Gemmell CG. Staphylococcal scalded skin syndrome. J Med Microbiol 1995; 43:318–327. 22. Cribier B, Piemont Y, Grosshans E. Staphylococcal scalded skin syndrome in adults: a clinical review illustrated with a case. J Am Acad Dermatol 1984; 30:319–324. 23. Brochers SL, Gomez EC, Isseroff RR. Generalized staphylococcal scalded skin syndrome in an anephric boy undergoing hemodialysis. Arch Dermatol 1984; 120:912–918. 24. Goldberg NS, Ahmed T, Robinson B, et al. Staphylococcal scalded skin syndrome mimicking acute graft-versus-host disease in a bone marrow transplant recipient. Arch Dermatol 1989; 125:85–87. 25. Lowy FD. Staphylococcus aureus infections. N Engl J Med 1998:520–532. 26. Markowitz N, Quinn EL, Saravolatz LD. Trimethoprim-sulfamethoxazole compared with vancomycin for the treatment of Staphylococcus aureus bacteremia. Ann Intern Med 1992; 117(5):390–398. 27. Anstead GM, Owens AD. Recent advances in the treatment of infections due to resistant Staphylococcus aureus. Curr Opin Infect Dis 2004; 17:549–555. 28. Manders SM. Toxin-mediated streptococcal and staphylococcal disease. J Am Acad Dermatol 1998; 39(3):383–398. 29. Centers for Disease Control and Prevention. Epidemiology Program Office, Division of Public Health surveillance and Informatics. Case definitions for infectious conditions under public health surveillance. Part 1. Case definitions for nationally notifiable infectious disease. Available at http://www.cdc.gov/epo/dphsi/casedef/notiable.htm (accessed June 11, 2005). 30. The Working Group on Severe Streptococcal Infections. Defining the group A streptococcal toxic shock syndrome: rationale and consensus definition. JAMA 1993; 269:390–391. 31. Chesney PJ, Davis JP, Purdy WK, et al. Clinical manifestation of toxic shock syndrome. JAMA 1981; 246:741–748. 32. Shands KN, Schmid GP, Dan BB, et al. Toxic shock syndrome in menstruating women. Association with tampon use and Staphylococcus aureus and clinical features in 52 cases. N Engl J Med 1980; 303:1436–1442. 33. Parsonnet J. Non menstrual toxic shock syndrome: new insights into diagnosis, pathogenesis, and treatment. Curr Clin Top Infect Dis 1996; 16:1–20. 34. Kain KC, Schulzer M, Chow AW. Clinical spectrum of nonmenstrual toxic shock syndrome (TSS): comparison with menstrual TSS by multivariate analyses. Clin Infect Dis 1993; 16:100–106. 35. Davis JP, Osterholm MT, Helms CM, et al. Tri-state toxic shock syndrome study. II. Clinical and laboratory findings. J Infect Dis 1982; 145:441–448. 36. Davis JP, Chesney PJ, Wand PJ, et al. Toxic-shock syndrome: epidemiologic features, recurrence, risk factors, and prevention. N Engl J Med 1980; 303:1429–1435. 37. Stolz SJ, Davis JP, Vergeront JM, et al. Development of serum antibody to toxic shock toxin among individuals with toxic shock syndrome in Wisconsin. J Infect Dis 1985; 151:883–889. 38. Stevens DL. Invasive group A streptococcus infections. Clin Infect Dis 1992; 14:2–13. 39. Johnson Dr, Stevens DL, Kaplan EL. Epidemiologic analysis of group A streptococcus serotypes associated with severe systemic infections, rheumatic fever, or uncomplicated pharyngitis. J Infect Dis 1992; 166:374–382. 40. Teglund AN, Thulin P, Gan BS. Evidence for superantigen involvement in severe group A streptococcal tissue infections. J Infect Dis 2001; 184:853–860.

Severe Skin and Soft Tissue Infections in the Critical Care Unit

343

41. Hackett SP, Stevens DL. Streptococcal toxic shock syndrome: synthesis of tumor necrosis factor and interleukin-1 by monocytes stimulated with pyrogenic exotoxin A and streptolysin O. J Infect Dis 1992; 165:879–885. 42. Fast DJ, Schlievert PM, Nelson RD. Toxin shock syndrome-associated staphylococcal and streptococcal pyrogenic toxins are potent inducers of tumor necrosis factor production. Infect Immun 1989; 57:291–294. 43. Saouda M, Wu W, Conran P, et al. Streptococcal pyrogenic exotoxin B enhances tissue damage initiated by other Streptococcus pyogenes products. J Infect Dis 2001; 184: 723–731. 44. Proft T, Sriskandan S, Yang L. Superantigens and streptococcal toxic shock syndrome. Emerg Infect Dis 2003; 9:1211–1218. 45. Forni AL, Kaplan EL, Schlievert PM, Roberts RB. Clinical and microbiological characteristics of severe group A streptococcus infections and streptococcal toxic shock syndrome. Clin Infect Dis 1995; 21(2):333–340. 46. Factor SH, Levine OS, Schwartz B, et al. Invasive group A streptococcal disease: risk factors for adults. Emerg Infect Dis 2003; 8:970–977. 47. McCormick JK, Yarwood JM, Schlievert PM. Toxic shock syndrome and bacterial superantigens: an update. Annu Rev Microbiol 2001; 55:77–104. 48. Darenberg J, So¨derquist B, Normark BH. Differences in potency of intravenous polyspecific immunoglobulin G against streptococcal and staphylococcal superantigens: implications for therapy of toxic shock syndrome. Clin Infect Dis 2004; 38:836–842. 49. Stevens DL. Dilemmas in the treatment of invasive streptococcus pyogenes infections. Clin Infect Dis 2003; 37:341–343. 50. Eagle H. Experimental approach to the problem of treatment failure with penicillin. I. Group A streptococcal infection in mice. Am J Med 1952; 13:389. 51. Stevens DL, Gibbons AE, Bergstrom R, et al. The Eagle effect revisited: efficacy of clindamycin, erythromycin, and penicillin in the treatment of streptococcal myositis. J Infect Dis 1988; 158:23–28. 52. Stevens DL, Yan S, Bryant AE. Penicillin-binding protein expression at different growth stages determines penicillin efficacy in vitro and in vivo: an explanation for the inoculum effect. J Infect Dis 1993; 167:1401–1405. 53. Stevens DL, Bryant AE, Yan S. Invasive group A streptococcal infection: new concepts in antibiotic treatment. Int J Antimicrob Agents 1994; 4:297–301. 54. Gemmell CG, Peterson PK, Schmeling D, et al. Potentiation of opsonization and phagocytosis of Streptococcus pyogenes following growth in the presence of clindamycin. J Clin Invest 1981; 67:1249–1256. 55. Kaul R, McGeer A, Norrby-Teglund A, et al. Intravenous immunoglobulin therapy for streptococcal toxic shock syndrome—a comparative observational study. Canadian Streptococcal study group. Clin Infect Dis 1999; 28:800–807. 56. Darenberg J, Ihendyane N, Sjolin J, et al. Intravenous immunoglobulin G therapy in streptococcal toxic shock syndrome: a European randomized, double blind, placebo controlled trial. Clin Infect Dis 2003; 37:333–340. 57. Norrby-Telgund A, Basma H, Anderson J, et al. Varying titres of neutralizing antibodies to streptococcal superantigens in different preparations of normal polyspecific immunoglobulin G (IVIG): implications for therapeutic efficacy. Clin Infect Dis 1998; 26: 631–638. 58. Jorup-Ronstrom C. Epidemiological and complicating features of erysipelas. Scand J Infect Dis 1986; 18:519–524. 59. Swartz MN. Cellulitis. N Engl J Med 2005; 350:904–912. 60. Swartz MN, Pasternack MS. Cellulitis and subcutaneous tissue infections. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practices of Infectious Disease. 6th ed. Philadelphia: Elsevier, Churchill Livingstone, 2005; 1:1172–1194.

344

Sharma and Saravolatz

61. Bisharat N, Agmon V, Finkelstein R, et al. Clinical, epidemiological, and microbiological features of Vibrio vulnificus biogroup 3 causing outbreaks of wound infection and bacteraemia in Israel. Israel Vibrio Study Group. Lancet 1999; 354:1421–1424. 62. Bonner JR, Coker AS, Berryman CR, et al. Spectrum of vibrio infections in a gulf coast community. Ann Intern Med 1983; 99:464–469. 63. Gold WL, Salit IE. Aeromonas hydrophilia infections of the skin and soft tissue: report of 11 cases and review. Clin Infect Dis 1993; 16:69–74. 64. Weinstein MR, Litt M, Kertesz Da, et al. Invasive infections due to a fish pathogen, Streptococcus iniae. S. iniae Study Group. N Engl J Med 1997; 337(9):589–594. 65. Capdevila O, Grau I, Vadillo M, et al. Bacteremic pneumococcal cellulitis compared with bacteremic cellulitis caused by Staphylococcus aureus and Streptococcus pyogenes. Eur J Clin Microbiol Infect Dis 2003; 22(6):337–341. 66. Porras MC, Martinez VC, Ruiz IM, et al. Acute cellulitis: an unusual manifestation of meningococcal disease. Scand J Infect Dis 2001; 33(1):56–59. 67. Brooke CJ, Riley TV. Erysipelothrix rhusiopathiae: bacteriology, epidemiology and clinical manifestations of an occupational pathogen. J Med Microbiol 1999; 48(9):789–799. 68. Talan DA, Citron DM, Abrahamian FM, et al. Bacteriologic analysis of infected dog and cat bites. N Engl J Med 1999; 340:85–92. 69. Moal GL, Landron C, Grollier G, et al. Meningitis due to Capnocytophaga canimorsus after receipt of a dog bite: case report and review of the literature. Clin Infect Dis 2003; 36:e42–e46. 70. Talan DA, Abrahamian FM, Moran GJ, et al. Clinical presentation and bacteriologic analysis of infected human bites in patients presenting to emergency departments. Clin Infect Dis 2003; 37:1481–1489. 71. Greene SL, Su WP, Muller SA. Ecthyma gangrenosum: report of clinical, histopathologic, and bacteriologic aspects of eight cases. J Am Acad Dermatol 1984; 11:781–787. 72. Huminer D, Siegman-Igra Y, Morduchowicz G, et al. Ecthyma gangrenosum without bacteremia. Report of six cases and review of the literature. Arch Intern Med 1987; 147(2): 299–301. 73. Sevinsky LD, Viecens C, Ballesteros DO, et al. Ecthyma gangrenosum: a cutaneous manifestation of Pseudomonas aeruginosa sepsis. J Am Acad Dermatol 1993; 29(1):104–106. 74. Inglesby TV, O’Toole, Henderson DA, et al. Anthrax as a biological weapon 2002: updated recommendation for management. JAMA 2002; 287:2236–2252. 75. Swartz MN. Recognition and treatment of Anthrax: an update. N Engl J Med 2001; 345:1621–1626. 76. Centers for Disease Control and Prevention update: investigation of bioterrorism resulted anthrax and interim guidelines for exposure management and antimicrobial therapy. Mor Mortal Wkly Rep 2001; 50:909–919. 77. Harrison OB, Robertson BD, Faust SN, et al. Analysis of pathogen host cell interactions in purpura fulminans: expression of capsule, type IV pilli, and Por A by Neisseria meningitides in vivo. Infect Immun 2002; 70:193–201. 78. Carpenter CT, Kaiser AB. Purpura fulminans in pneumococcal sepsis: case report and review. Scand J Infect Dis 1997; 29(5):479–483. 79. Kravitz GR, Dries DJ, Peterson ML, et al. Purpura fulminans due to Staphylococcus aureus. Clin Infect Dis 2005; 40:941–947. 80. Seal DV. Necrotizing fasciitis. Curr Opin Infect Dis 2001; 14:127–132. 81. Kuncir EJ, Tillou A, St. Hill CR, et al. Necrotizing soft tissue infections. Emerg Med Clin N Am 2003; 21:1075–1087. 82. Nicholas RL, Florman S. Clinical presentations of soft-tissue infections and surgical site infections. Clin Infect Dis 2001; 33:S84–S93. 83. McHenry CR, Pitrowski JJ, Petrinic D, et al. Determinants of mortality in necrotizing soft tissue infections. Ann Surg 1995; 221:558–563. 84. Giuliano A, Lewis F Jr., Hardy K, et al. Bacteriology of necrotizing fasciitis. Am J Surg 1977; 134:52–57.

Severe Skin and Soft Tissue Infections in the Critical Care Unit

345

85. Sharma M, Khatib R, Fakih M. Clinical characteristics of necrotizing fasciitis caused by Group G streptococcus: case report and review of the literature. Scand J Infect Dis 2002; 34(6):468–471. 86. Patino JF, Castro D. Necrotizing lesions of the soft tissue: a review. World J Surg 1991; 15:235–239. 87. Chapnik EK, Abter EI. Necrotizing soft tissue infections. Infect Dis Clin North Am 1996; 10:835–855. 88. Cunningham JD, Silver L, Rudikoff D. Necrotizing fasciitis. A plea for early diagnosis and treatment. Mt Sinai J Med 2001; 68:253–261. 89. Wong CH, Wang YS. The diagnosis of necrotizing fasciitis. Curr Opin Infect Dis 2005; 18:101–106. 90. Kaul R, McGeer A, Low DE, Green K, Schwartz B. Population-based surveillance for group A streptococcal necrotizing fasciitis: clinical features, prognostic indicators, and microbiologic analysis of seventy-seven cases. Ontario Group A Streptococcal Study. Am J Med 1997; 103(1):18–24. 91. Anderson TJ, Green SD, Childers BJ. Massive soft tissue injury: diagnosis and management of necrotizing fasciitis and purpura fulminans. Plast Reconstr Surg 2001; 107: 1025–1035. 92. Jones RB, Hirschmann JV, Brown GS, et al. Fournier’s syndrome: necrotizing subcutaneous infection of the male genitalia. J Urol 1979; 122; 279–282. 93. Nickels JC, Morales A. Necrotizing fasciitis of the male genitalia (Fournier’s gangrene). Can Med Assoc J 1983; 129:445–448. 94. Cline KA, Turnbull TL. Clostridial myonecrosis. Ann Emerg Med 1985; 14:459–466. 95. Gorbach SL, Thadepalli H. Isolation of clostridium in human infections: evaluation of 114 cases. J Infect Dis 1975; 131:S81–S85. 96. Bryant AE, Chen RYZ, Nagata Y. Clostridial gas gangrene. I. Cellular and molecular mechanisms of microvascular dysfunction induced by exotoxins of Clostridium perfringens. J Infect Dis 2000; 182:799–780. 97. Stevens DL, Bryant AE. The role of clostridial toxins in the pathogenesis of gas gangrene. Clin Infect Dis 2002; 35:S93–S100. 98. Stephans MB. Gas gangrene. Potential for hyperbaric oxygen therapy. Postgrad Med 1996; 99:217–220. 99. Tibbles PM, Edelsberg JS. Hyperbaric-oxygen therapy. N Engl J Med 1996; 334: 1642–1648. 100. Stevens DL, Musher DM, Watson DA, et al. Spontaneous nontraumatic gangrene due to Clostridium septicum. Rev Infect Dis 1990; 12:286–296. 101. DiNubile MJ, Lipsky BA. Complicated infections of skin and skin structures: when the infection is more than skin deep. J Antimicrob Chemother 2004; 53(S2):ii37–ii50. 102. Pasternack MS, Swartz MN. Myositis. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practices of Infectious Disease. 6th ed. Philadelphia: Elsevier, Churchill Livingstone, 2005; 1:1194–1204. 103. Lipsky BA, Pecoraro RE, Wheat JL. The diabetic foot: soft tissue and bone infection. Infect Dis Clin North Am 1990; 4:409–432. 104. Wheat LJ, Allen SD, Henry M, et al. Diabetic foot infections: bacteriologic analysis. Arch Intern Med 1986; 146:1935:40. 105. Sapico FL, Witte JL, Canawati HN, et al. The infected foot of the diabetic patient: quantitative microbiology and analysis of clinical features. Rev Infect Dis 1984; 6(suppl 1): 171–176. 106. Murphy EL, DeVita D, Liu H, et al. Risk factors for skin and soft-tissue abscesses among Injection drug users: a case control study. Clin Infect Dis 2001; 33:35–40. 107. Chen JL, Fullerton KE, Flynn NM. Necrotizing fasciitis associated with injection drug use. Clin Infect Dis 2001; 33:6–15. 108. Ebright JR, Pieper B. Skin and soft tissue infections in injection drug users. Infect Dis Clin North Am 2002; 16:697–712.

346

Sharma and Saravolatz

109. Levin MJ, Gardner P, Waldvogel FA. Tropical pyomyositis: an unusual infection due to Staphylococcus aureus. N Engl J Med 1971; 284:196–198. 110. Crum NF. Bacterial pyomyositis in the United States. Am J Med 2004; 117:420–428. 111. Miralles GD, Bregman Z. Necrotizing pyomyositis caused by Mycobacterium avium complex in a patient with AIDS. Clin Infect Dis 1994; 18:833–834. 112. Sharma M, Khatib R, Fakih M. Cryptococcus neoformans myositis without dissemination. Scand J Infect Dis 2002; 34(11):858–859. 113. Fridkin SK, Hageman JC, Morrison M, et al. Methicillin resistant staphylococcus aureus disease in three communities. N Engl J Med 2005; 352:1436–1444. 114. Johnson LB, Saravolatz LD. Community-acquired MRSA: current epidemiology and management issues. Infect Med 2005; 22:16–20. 115. Miller LG, Perdeau-Remington, Reig Gunter, et al. Necrotizing fasciitis caused by community associated methicillin resistant Staphylococcus aureus in Los Angeles. N Engl J Med 2005; 352:1445–1453.

17 Infections in Patients on Steroids in the Critical Care Unit John N. Sheagren Department of Internal Medicine, Advocate Illinois Masonic Medical Center, Chicago, Illinois, U.S.A.

INTRODUCTION For over half a century, steroids (corticosteroids and glucocorticoids) have been used to suppress inflammatory, autoimmune, and other immunologic processes (1–3). Their benefits are unquestioned; however, their costs are high. When needed, their multifactorial enhancing and suppressing molecular and cellular effects on metabolic systems lead to hypertension, electrolyte abnormalities, and mineral disorders, especially osteoporosis (1–3). Their impacts on inflammatory systems impair tissue integrity and repair and result in a variety of infectious complications (4), the focus of this chapter. While steroids still need to be used in nearly every specialty of medicine, clinicians are constantly searching for ‘‘steroid sparing,’’ immunosuppressive/ anti-inflammatory treatments for the multitudes of inflammatory and autoimmune disorders suffered by their patients. This chapter will provide an historical overview of the relationship between steroids and infections, how the host’s defense system is organized to defend against microbial invasion by specific organisms, how steroids alter each aspect of the inflammatory system generally to increase the likelihood of each such infection, and, paradoxically, how steroids can actually be used to treat some specific infectious diseases. Several years ago, all these topics were reviewed in great detail (5); and most of the data and concepts provided in that review are still relevant today.

HISTORICAL ASPECTS OF STEROIDS AND INFECTIONS The historical relationships between the function of the adrenal gland and infection were masterfully summarized in a review article in 1953 by Kass and Finland (6). They pointed out that Dr. Harvey Cushing in the early 1900s noted that adrenalectomy 347

348

Sheagren

(hypoadrenalism) was associated with increased susceptibility to and mortality from infection. Physiologic replacement doses of an adrenal extract, containing the adrenal cortical hormones, corrected the problem of increased mortality from infection in experimental animals. The evolution through the late 1800s and early 1900s of the understanding of the adrenal cortex as compared to the adrenal medulla lead to the observation that the life of experimental animals could be prolonged by the administration of adrenal cortical but not adrenal medullary (epinephrine hormone) administration. Not only was the life of the adrenalectomized animal extended with adrenal cortical extracts, but such extracts also corrected the increased susceptibility to infection as well as the other metabolic and hemodynamic abnormalities which they had described. In 1912, Dr. Harvey Cushing was the first to define the disease produced by adrenal cortical steroid excess. It was not for another 20 years, however, that he wrote the definitive paper on the syndrome, which subsequently came to be called ‘‘Cushing’s Syndrome.’’ While such patients were hypertensive and experienced a variety of fluid and electrolyte abnormalities, Dr. Cushing noted that infection was the primary cause of death in about half (6). In the late 1940s, Hench et al. (7) at the Mayo Clinic described the first clinical uses of supraphysiologic doses of cortisone on a variety of systemic inflammatory diseases, especially rheumatoid arthritis. Not long afterward, case reports appeared describing a wide variety of infectious complications occurring in patients on such doses of steroids. Kass went on to perform a series of elegant laboratory studies confirming that steroid therapy produced a dose-related increase in susceptibility to a variety of infectious agents in experimental animals (8).

HOST DEFENSES/IMMUNITY AGAINST INFECTIONS An excellent review of the actions of steroids on immunity has recently been published (9). That review noted and summarized a variety of studies over the past half century that at least have partially clarified the multiple actions of steroids on the immune system. The complexity of such effects is enormous: In some situations, steroids up regulate and enhance immunologic functions, while in others, they down regulate and/or suppress immunity. Franchimont (9) outlines in some detail the molecular, cellular, and pharmacologic properties of steroids and pointed out that steroids exert both negative and positive effects on various limbs and components of the immune response. While modulating genes involved in the priming of the innate (nonspecific) immune response, they suppress cellular [T helper-1 (Th-1)] immunity and promote humoral [T helper 2 (Th-2)] immunity in the adaptive (specific or acquired) immune response. Franchimont goes on to suggest the ex vivo therapeutic use of steroids might represent a way positively to modulate cellular responses in autoimmune diseases while avoiding long-term systemic steroid side effects. In my earlier review of this topic (5), I proposed a simple organization of the immune system in order to permit a better understanding of how each component functions to defend the host against microbial invaders, as well as in a more focused fashion to help explain how steroids affect each limb of immunity. Table 1 outlines an organizational structure of the immune system, useful in understanding the host defense systems against infections. Briefly, the immune system is broken down into the nonspecific (innate) and the specific (acquired or adaptive) immune systems.

Infections in Patients on Steroids in the Critical Care Unit

349

Table 1 A Simplified Organization of the Host Defense System Nonspecific (innate) immunity Barrier systems: skin, mucous membranes, cilia, and mucus Complement system (via the alternative pathway) WBCs: neutrophils, macrophages, and eosinophils Specific (acquired or adaptive) immunity Macrophages, including dendritic (antigen processing) cells Cellular immune response: mediated by Th-1 lymphocytes, which release cytokines to activate surrounding macrophages which then become more efficient microbial killers Humoral immune response: mediated by Th-2 lymphocytes, which subsequently activate surrounding B lymphocytes to evolve into plasma cells which produce the various classes of immunoglobulins/antibodies. Antibodies either opsonize bacteria to permit phagocytosis or directly lyse and kill bacteria with the help of the complement system (this time via the ‘‘classical’’ pathway) Abbreviations: Th-1, T helper 1; Th-2, T helper 2; WBCs, white blood cells.

The Nonspecific (Innate) Immune System The first lines of defense against microbial invaders are the skin and mucous membranes, breaks in which, for example, create entry points for colonizing acute invasive bacteria, mostly staphylococci and streptococci, to enter the body. Once in the subcutaneous or submucosal tissues, these microbes replicate and initiate activation (via a variety of their cell materials, toxins, and enzymes) of the complement system, at this initial point through its ‘‘alternative’’ pathway (10). Complement components not only enhance local vasoactive and mediator contributions to the local inflammatory response, but generate chemotactic fragments which draw in surrounding leukocytes. The white blood cells (WBCs) are the most important nonspecific participants in this initial phase of host defense, for complement’s alternative pathway does not itself directly participate in bacterial killing. Because all organisms, whether opsonized or not, activate the alternative limb of the compliment pathway, it is the response of WBCs, predominantly the neutrophils and macrophages, which are the most important host defense elements in this initial, nonspecific phase of the host defenses. The vast majority of microbial invaders are stopped in their tracks by the combination of the barrier systems and the complement assembled WBCs. The Specific (Acquired or Adaptive) Immune System Macrophages (including dendritic cells) are crucial to antigen processing and initiate both the cellular (Th-1) lymphocyte activation and response and the humoral (Th-2) immune response. Further, macrophages are also effector cells in the fully developed cellular immune response, responsible, when activated, for intracellular killing of a variety of important pathogens. Th-1 lymphocytes reacting to specific antigen stimulation release cytokines, which subsequently assemble and activate available surrounding macrophages to become highly efficient killers of facultative intracellular microbes; such organisms can survive ingestion in unactivated macrophages and go on to cause further tissue damage (11). The humoral immune response has as its primary role the production of immunoglobulins, most importantly antibodies of the immunoglobulin M (IgM) and immunoglobulin G (IgG) classes (12). These antibodies are either directly bactericidal [along with complement (10), this time via the ‘‘classical’’ pathway] or important

350

Sheagren

opsonins leading to more efficient phagocytosis and cellular killing, especially of a variety of encapsulated microbes (the prototypic organism being Streptococcus pneumoniae). Opsonins probably also help in the defense against a variety of gram-negative organisms, especially neisseria and certain other gram-negative rods.

MICROBIAL DEFENSE BY THE VARIOUS COMPONENTS OF THE IMMUNE SYSTEMS While the preceding paragraphs have discussed the components of the host defense/ immune systems and alluded to some of the microbes dealt with by each component of immunity, the next few paragraphs will focus on the specific microbes expected when there are deficiencies in those host defense systems (Table 2). Infections Expected in Deficiencies of the Nonspecific/Innate Immune System The barrier systems are most important in protecting against invasion by the acute pyogenic bacteria such as the beta-hemolytic streptococci and Staphylococcus aureus. The classic example of such defects is the patient with exfoliative types of dermatitis (such as an exfoliating drug eruption or Stevens/Johnson Syndrome) or the extensively burned patient (13). The sequence of events in such patients is usually the following: first occurs colonization and then infection with gram-positive cocci (streptococci and staphylococci), which are generally antibiotic susceptible and therefore treatable. Next come the colonization and then infection with gram-negative bacilli such as the coliform bacteria, Klebsiella, and Pseudomonas aeruginosa. Intense antibacterial antibiotic therapy may again stabilize such patients for a while, but then overgrowth with a variety of yeasts, such as Candida albicans, or the other candida species creates an extremely difficult clinical situation. Also, burn wound colonization and infection with a variety of viruses sometimes occur (such as the herpes viruses or other DNA viruses). Deficiencies in the complement system may Table 2 Microbes Defended Against by the Various Components of the Host Defense/ Immune Systems Immune system component Nonspecific/innate immune system (barrier systems, complement via the ‘‘alternative’’ pathway, and leukocytes, mainly the neutrophil) Specific (acquired or adaptive) immunity: Cellular immunity (Th-1, lymphocytes, and macrophages) Humoral immunity (Th-2 lymphocytes, plasma cells, and complement via the ‘‘classical’’ pathway)

Defends against Acute pyogenic bacteria (streptococci, staphylococci, some gram negatives such as neisseria, etc.)

The group of microbes collectively known ‘‘facultative intracellular pathogens’’ (see text) IgM and IgG antibodies protect primarily encapsulated, pyogenic bacteria (classically, Streptococcus pneumoniae, Hemophilus influenzae, and Neisseria; may also help against some other gram negative rods

Abbreviations: Th-1, T helper 1; Th-2, T helper 2; IgM, immunoglobulin M; IgG, immunoglobulin G.

Infections in Patients on Steroids in the Critical Care Unit

351

lead also to infections with acute pyogenes and seem particularly to increase susceptibility to neisseria (14). Finally, defects in numbers or function of neutrophils classically lead to infections with streptococci and staphylococci and, after repeated courses of antibiotics, to a variety of gram-negative bacilli, and then yeasts. Specific (Acquired or Adaptive) Immunity Cellular Immunity As noted earlier, the cellular immune system is composed of Th-1 lymphocytes, the cytokines they produce, and the macrophages which on the afferent side process antigen to initiate the cellular immune response and on the efferent side become activated by the Th-1 lymphocytokines to become the ultimate killer cells of cellular immunity. As shown in Table 2, the cellular immune system defends against the group of organisms collectively known as the ‘‘facultative intracellular pathogens’’ (FIPs), microorganisms that generally survive inside nonimmune host macrophages (15). These types of microbes may be effectively dealt with when they are ingested by cytokine ‘‘activated’’ macrophages, i.e., those whose microbicidal systems have been turned on by cytokines released by specifically sensitized Th-1 lymphocytes (11,16). Thus, in nonimmune hosts, FIPs may live symbiotically within macrophages, deriving, in fact, some of their metabolic needs from the host cells themselves. However, once the immune response has been generated by antigen processing cells, which program subsets of Th-1 lymphocytes against antigens peculiar to each specific invading FIP, such organisms are either killed, or at least held in check, for prolonged periods of time. Those organisms, lying dormant but still alive within host tissues, mostly within activated macrophages (which pathologically are recognized as epitheloid cells or giant cells combining to form granulomas), may, when the cellular immune responses have become suppressed, begin to grow and cause a ‘‘reactivated’’ infection, proceeding on to cause local and/or systemic host damage. The major causes of the suppression of the cellular immune system are prolonged periods of exogenous stress, usually accompanied by malnutrition, certain specific viral infections, especially the human immunodeficiency virus, and systemic immunosuppressive agents, particularly steroids. The types of microbes known as FIPs contained by cellular immunity and which are most important in infecting humans (as noted in Table 2) are several types of bacteria (listeria, salmonellae, mycobacteria, and nocardia) and fungi (cryptococcus, histoplasmosis, coccidioidomycosis, and the yeasts, especially candida species). Several parasites are also primarily controlled by the cellular immune system (toxoplasmosis, Pneumocystis carinii, leishmania, cryptosporidium, and strongyloids). Humoral Immunity The humoral immune system consists of Th-2 lymphocytes and the B cells, which are activated and programmed to evolve into plasma cells producing the various classes of immunoglobulins. These immunoglobulins/antibodies are important in defending against a variety of pathogens. As noted earlier, the humoral/antibody mediated immune system also is aided by the complement system, now via the ‘‘classical’’ pathway, and is important in both opsonization and direct bactericidal activity (10,12). Particularly important in host defense are the IgM and IgG antibody classes. Such antibodies help protect the host primarily against the encapsulated, pyogenic bacteria, the classical organisms being S. pneumoniae, Hemophilus influenzae, and Neisseria (12).

352

Sheagren

Antibody mediated defenses may also help against some other gram-negative rods such as the coliform bacteria and klebsiella (Table 2).

MAJOR EFFECTS OF STEROIDS ON IMMUNITY It is important for the reader to be aware that different doses of steroids (glucocorticoids) are used to achieve different therapeutic results in patients; further, different steroid doses have quite different effects on host inflammation and immunity (Table 3). I have arbitrarily classified these dose effects into the following ranges: physiologic, pharmacologic, and suprapharmacologic doses of steroids. Physiologic Doses of Steroids Physiologic doses of steroids are those in the daily range from 20 to 30 mg of hydrocortisone, the amount that the unstressed adrenal cortex produces over 24 hours (15). Thus, if a patient is on from 5 to 10 mg of prednisone for an underlying inflammatory disease, such a dose range (equivalent to 20–40 mg of hydrocortisone) represents essentially physiologic replacement of adrenal cortical function. Individuals on such doses have a relatively intact immune system, although a variety of long-term metabolic effects do occur (especially bone loss and osteoporosis). Such patients are not at a substantially increased risk of infectious complications. One needs to be aware, however, that should such a patient encounter the increased stress of trauma or infection, they then need to be treated with increased stress doses of hydrocortisone, in the range from 100 to 200 mg (20–40 mg of prednisone) daily. Again, patients treated properly during such periods of stress seem to be at no increased risk of infection, unless the period of stress and the administration of the aforementioned doses of glucocorticoids are prolonged beyond a few days (15). Pharmacologic Doses of Steroids These doses of prednisone (hydrocortisone is rarely used in such situations) are in the range of 1 mg/kg (40–100 mg) per day, doses typically used to treat a variety of acute inflammatory/autoimmune conditions. Such doses of prednisone are equivalent to 200 to 500 mg of hydrocortisone daily and when continued for more than four to six weeks, begin to produce the classical changes of Cushing’s Syndrome with fat redistribution (moon facies, buffalo hump, and truncal obesity) and thinning of the skin, including the presence of striae and increasing bruisability. Further, these doses of steroids effectively block the access of acute inflammatory cells, especially neutrophils and macrophages, to foci of inflammation and infection, resulting in both reduced symptoms of the underlying inflammatory disease and reduced evidence of an infection as it progresses (16,17). Infections expected under such conditions are those caused by the acute pyogenic bacteria (again, streptococci and staphylococci). Also, rapid colonization and usually only superficial invasion occur with yeasts (C. albicans and other candida species), resulting in oral, pharyngeal, vaginal, and intertriginous infections (16). The cellular (Th-1) immune system becomes suppressed after two to four weeks of such doses of steroids, resulting in impaired access of macrophages to inflammatory and infectious sites and decreased cytokine activation of macrophages by already programmed Th-1 lymphocytes (18). In such patients, positive delayed type

None (unless prolonged)

None (unless prolonged)

Mostly infections with acute pyogenes, then Massive suppression of nonspecific/innate, (if such doses are continued) with gramcellular and even humoral immunity; effects negative rods, then with facultative present within 48 hrs; these negative immune intracellular pathogens effects reverse quickly when such doses are discontinued Effects on the immune systems of pulse doses not None or acute pyogenic bacterial infection well studied; probably minimal at 3–5 days; (at 5–7 days) some effects probable at 7 days and beyond

Acute pyogenic bacteria (staphylococci, Nonspecific/innate immunity impairs streptococci, superficial yeast infections) inflammatory cell (esp. neutrophil) access to inflammatory/infections foci; increases fragility and therefore likelihood of breaks in the barrier systems (skin/mucous membranes) Facultative intracellular pathogens (e.g., Cellular (Th-1) immunity becomes markedly mycobacteria, systemic mycoses, listeria, suppressed within 14–28 days; results in salmonella, chlamydia, rickettsia, impaired access of macrophages (MP) to trypanosoma, leishmania, and strongyloids) inflammatory/ infectious sites and deceases MP activation; converts positive delayed type skin tests (e.g., tuberculin) to negative Humoral (Th-2) immunity: These doses of None in particular; however, note that even steroids produce little effect; may even opsonized bacteria require the neutrophil for enhance some antibody responses killing!

None

Infections expected

None

Effect on immunity

a Trauma, infection, metabolic hyper- or hypoactivity (e.g., hypo- or hyperthyroidism, pregnancy, severe fluid/electrolyte abnormalities, etc.). Abbreviations: MP, macrophages; Th-1, T helper 1; Th-2, T helper 2.

Usually given in ‘‘pulse’’ doses over a 3 to 5 to 7–day period

‘‘Suprapharmacologic’’ dose range: 5000–10,000 mg of hydrocortisone (1–2 g of prednisone or equivalent) per day

Physiologic: Nonstressed dose range state: 20–30 mg of hydrocortisone (5–7.5 mg of prednisone or equivalent) Stressed statea: 100–200 mg of hydrocortisone (20–40 mg of prednisone) per day Pharmacologic dose range: 200–500 mg of hydrocortisone (40–100 mg of prednisone or equivalent)

Steroid doses/day

Table 3 Approximate Dose–Response Effects of Steroids on Host Defense/Immune Systems Infections in Patients on Steroids in the Critical Care Unit 353

354

Sheagren

hypersensitivity skin tests turn negative (19,20), and latent infections with FIP may become reactivated (21) and disseminate (e.g., those caused by tuberculosis, histoplasmosis, leishmania, strongyloidiasis, etc.). The humoral (Th-2) immune system is not substantially affected even by these pharmacologic doses of steroids (22,23). It is important to realize, of course, that even though S. pneumoniae may be effectively opsonized, the neutrophil is still required in order to ingest and kill the organism. Thus, it is safe to say that all types of acute pyogenic infections, both with the nonencapsulated and encapsulated bacteria, occur more frequently in patients on pharmacologic doses of glucocorticoids. Superpharmacologic (‘‘Mega-’’) Doses of Steroids Here, I refer to doses of steroids in the range from 1 to 2 g of prednisone (30–60 mg/kg, equivalent to doses of 5000–10,000 mg of hydrocortisone) given in some trials every six to eight hours for several days. When dosed in that fashion, secondary infections occur more frequently than in control subjects (24,25). At present, such doses are usually administered as a single daily dose (so called ‘‘pulse dose’’ regimens) over a three- to five- to seven-day period acutely to suppress all immunologically mediated inflammatory processes and/or to provide a variety of dermatological or anticancer effects (26,27). Sustained administration of 2 g of prednisone or its equivalent over a 48-hour period has been shown to suppress all limbs of the immune system and, if continued, will result in an increased rate of infections (24,25). The ‘‘pulse dose’’ approach of single daily doses of up to 2 g of methylprednisolone per day for about a week to patients without underlying systemic diseases is relatively free of infectious side effects (28–30). As the period of therapy extends beyond five days, the frequency of infectious complications will undoubtedly increase. Also, the rate of infection is increased even with pulse dose therapy in patients with an underlying systematic disease such as systemic lupus erythematosus (30).

TYPES OF INFECTIONS EXPECTED IN PATIENTS ON STEROIDS From the preceding discussion, it should become clear that low, physiologic doses of steroids not only are necessary for patients with inadequate adrenal responses effectively to cope with and survive infections, but that such doses have little if any effect on any limb of the immune system. However, when one initiates physiologic ‘‘stress’’ doses of hydrocortisone (100–200 mg/day, usually given as 50 mg every six to eight hours), one begins to approach the pharmacologic dose range; such doses should be minimized to the period of extreme stress and then promptly and steadily reduced as the patient stabilizes. In fact, it is this author’s practice not to use more than 150 mg of hydrocortisone (37.5 mg of prednisone) daily even in patients in highly stressful metabolic/traumatic/infectious situations in order to decrease the likelihood of infections and other metabolic side effects. When patients are started on pharmacologic doses of steroids (Table 3), immunosuppression occurs fairly rapidly (3,31). In fact, within 12 to 24 hours, circulating leukocytes are affected, and an elevation of the WBC count occurs (2). The elevation of the WBC after steroid administration, especially neutrophils, is accompanied by a rapid decrease of eosinophils (32). The elevation of the neutrophil count is due to a number of factors, including enhanced proliferation and expansion of the neutrophil pool, decreased attachment of neutrophils to endothelial cells (decreased margination,

Infections in Patients on Steroids in the Critical Care Unit

355

which increases the pool of circulating neutrophils), and inhibition of granulocyte colony stimulating factor, which enhances the proliferation and expansion of the neutrophil pool. Monocytes too are ‘‘demarginated’’ and therefore also have deceased access to peripheral sites of inflammation/infection. High doses of steroids negatively affect both levels and function of the complement system itself (33,34). Pharmacologic doses of steroids must be continued for some weeks before impairment of barrier systems (skin, mucous membrane, etc.) is observed. Even a single superpharmacologic dose of steroids (1–2 g of prednisone or its equivalent) alters neutrophil, macrophage, and lymphocyte traffic from bone marrow to blood stream to lymphoid tissues as well as to peripheral foci of inflammation and infection. However, such effects are short lived; and as noted earlier, it seems that up to five daily boluses of such superphysiologic amounts of glucocorticoid are well tolerated, not producing a substantial increase in the propensity towards infection. Pharmacologic doses of steroids for more than four to eight weeks are probably required to substantially impair cellular immunity. Suppression occurs by inhibiting Th-1 lymphocyte-mediated macrophage activation, and disrupting epithelioid cell and giant cell (granuloma) formation, all of which could result in the reactivation of latent facultative infections (tuberculosis and histoplasmosis being the prototypes) (3).

STEROID TREATMENT OF INFECTION While this chapter has focused primarily on the adverse effects of steroid administration on the host resistance to infections, steroids also have clearly defined indications for use in specific infectious situations. The Infectious Diseases Society of America (IDSA) published guidelines for the systemic use of steroids in the treatment of certain infectious conditions in a well-done article in 1992 (35). A working group of the IDSA classified the strength of data to support recommendations for or against the use of steroid therapy and commented on the quality of evidence available to support a given recommendation. The following infections were felt to have ‘‘good evidence’’ to support a ‘‘recommendation for use’’ of pharmacologic doses of steroids: (i) typhoid fever resulting in critical illness (patient in shock) and (ii) tuberculous pericarditis. Those diseases in which there was ‘‘moderate evidence to support a recommendation for use’’ of pharmacologic doses of steroids included the following: (i) tetanus; (ii) tuberculous meningitis; and (iii) Epstein–Barr virus infection with impending airway obstruction. All the rest of the infectious situations for which steroids have been used in the past either with some data available or which were being used without any supportive evidence were felt to either have ‘‘poor evidence to support a recommendation for or against,’’ ‘‘moderate evidence to support a recommendation against’’, or ‘‘good evidence to support a recommendation against’’ steroid use. At the time of publication in 1992, the following illnesses fell into one or the other of those latter categories: septic shock syndrome, tuberculosis with severe constitutional syndromes, Herpes zoster, Epstein–Barr viral infection (including hepatitis, pericarditis, and encephalitis), hemorrhagic fever, trichinosis, and Kawasaki disease (35). Recently, well-done studies have shown the value of dexamethasone administration in both children (36) and adults (37) with meningitis. It is now recommended that such patients receive dexamethasone with or before the first dose of antibiotic (38,39).

356

Sheagren

Years ago, suprapharmacologic doses of steroids were regularly administered to patients in septic shock; however, controlled trials showed that such treatment did not work (24,25,40). More recently, data has become available in septic shock patients from reasonably well-done studies, which have resulted in a change in the recommendation for steroid use in patients suffering septic shock. A recent meta-analysis (41) and an accompanying editorial (42) put forth the following recommendations: low dose (physiologic/stress doses) steroids should be administered to all patients in septic shock; however, such doses should be continued only in patients with proven adrenal insufficiency. The dose of steroids referred to in the series of meta-analyzed studies is hydrocortisone 50 mg intravenously four times daily (total of 200 mg/day). At the time of steroid administration, a determination of the state of actual or relative adrenal insufficiency should be made; and the aforesaid dose of steroid should be continued beyond an initial two- to three-day period when the results of such testing are available. Only patients with proven adrenal insufficiency should be continued on such doses of hydrocortisone. Alternate-Day Steroid Therapy Every patient who requires long-term, pharmacologic doses of a glucocorticoid in order to control a serious inflammatory disorder should have regular attempts made to switch the daily regimen to one where the steroid dose is administered every other day in the morning (3,31). This ‘‘alternative-day’’ dosing schedule reduces the Cushingoid side effects, including reducing the risk of infection (20). For example, a patient requiring 60 mg of prednisone daily to control an autoimmune disease should always be worked, first, toward being dosed once daily in the morning. When stable, such patient should have the every-other day dose slowly (e.g., in 5 mg increments every one to two weeks) but surely reduced. If the patient begins to ‘‘flare’’ with recurrent symptoms of the underlying disease on the ‘‘off’’ day (the day the dose is being reduced), one should pause in terms of further reductions and try adding a nonsteroidal anti-inflammatory agent. In some patients, the symptoms are due to relative adrenal insufficiency rather than the underlying disease, in which case, a pause in further off-day dose reductions will lead to a gradual amelioration of the symptoms (tiredness, muscle aches, low grade fever, etc.). One may even need to increase the ‘‘on-day’’ dose for a brief period before attempting again to reduce the off-day dosage. Infectious complications in patients able to be transferred to an alternative-day steroid dosing regimen are substantially reduced; for example, delayed hypersensitivity type skin tests are preserved in such patients (20), and therefore underlying Th-1 (cellular immunity) is also probably intact. Such patients also seem quite able to withstand and respond well to the usual types of viral and bacterial infections (20,31).

CONCLUSIONS Steroids have major inhibitory effects on the nonspecific/innate immune system substantially impairing acute inflammatory cell (mainly neutrophil) access to early sites of infection; thus, not only are the rates of most infections increased in patients on steroids but the accompanying local and systemic signs and symptoms of infections are decreased, making early diagnosis of such episodes difficult. Most infections are caused by the acute pyogenic bacteria, the streptococci and staphylococci. Also

Infections in Patients on Steroids in the Critical Care Unit

357

adversely affected is the cellular (Th-1) immune system responsible for reacting to and controlling FIP (e.g., the tubercle bacillus); thus, progression and reactivation of such infections, though uncommon, may occur in patients on steroids. Minimal effects are produced by steroids on the humoral (Th-2) immune system. The awareness of the selective actions of steroids on host immunity to infections should assist clinicians in the earlier recognition and treatment of such potentially devastating complications.

REFERENCES 1. Pensabeni-Jasper T, Panush RS. Review: corticosteroid usage: observations at a community hospital. Am J Med Sci 1996; 311(5):234–239. 2. Boumpas DT, Chrousos GP, Wilder RL, Cupps TR, Balow JE. Glucocorticoid therapy for immune-mediated diseases: basic and clinical correlates. Ann Intern Med 1993; 119(12): 1198–1208. 3. Kehrl JH, Fauci AS. The clinical use of glucocorticoids. Ann Allergy 1983; 50(1):2–8. 4. Frey FJ, Speck RF. Glucocorticoids and infection. Schweiz Med Wochenschr 1992; 122(5): 137–146. 5. Sheagren JN. Glucocorticoid action: infectious diseases. In: Schleimer RP, Claman HN, Oronsky AL, eds. Anti-inflammatory Steroid Action: Basic and Clinical Aspects. Orlando, FL: Academic Press, 1989:525–543. 6. Kass EH, Finland M. Adrenocortical hormones in infection and immunity. Annu Rev Microbiol 1953; 7:361–388. 7. Hench PS, Kendall EC, Slocumb CH, Polley HF. Effects of cortisone acetate and pituitary ACTH on rheumatoid arthritis, rheumatic fever and certain other conditions. Arch Med Interna 1950; 85(4):545–666. 8. Kass EH. Hormones and host resistance to infection. Bacteriol Rev 1960; 24:177–185. 9. Franchimont D. Overview of the actions of glucocorticoids on the immune response: a good model to characterize new pathways of immunosuppression for new treatment strategies. Ann N Y Acad Sci 2004; 1024:124–137. 10. Gasque P. Complement: a unique innate immune sensor for danger signals. Mol Immunol 2004; 41(11):1089–1098. 11. Campbell PA. T cell involvement in resistance to facultative intracellular pathogens of the lung. Chest 1993; 103(suppl 2):113S–115S. 12. Ballow M. Primary immunodeficiency disorders: antibody deficiency. J Allergy Clin Immunol 2002; 109(4):581–591. 13. Pruitt BA Jr, McManus AT, Kim SH, Goodwin CW. Burn wound infections: current status. World J Surg 1998; 22(2):135–145. 14. O’Neil KM. Complement deficiency. Clin Rev Allergy Immunol 2000; 19:83–108. 15. Harty JT, Tvinnereim AR, White DW. CD8þ T cell effector mechanisms in resistance to infection. Annu Rev Immunol 2000; 18:275–308; van de Vosse E, Hoeve MA, Ottenhoff TH. Human genetics of intracellular infectious diseases: molecular and cellular immunity against mycobacteria and salmonellae. Lancet Infect Dis 2004; 4(12):739–749. 16. Holland EG, Taylor AT. Glucocorticoids in clinical practice. J Fam Pract 1991; 32(5): 512–519; Stuck AE, Minder CE, Frey FJ. Risk of infectious complications in patients taking glucocorticosteroids. Rev Infect Dis 1989; 11(6):954–963. 17. Ginzler E, Diamond H, Kaplan D, Weiner M, Schlesinger M, Seleznick M. Computer analysis of factors influencing frequency of infection in systemic lupus erythematosus. Arthritis Rheum 1978; 21(1):37–44. 18. Scheinman RI, Cogswell PC, Lofquist AK, Baldwin AS Jr. Role of transcriptional activation of I kappa B alpha in mediation of immunosuppression by glucocorticoids. Science 1995; 270(5234):283–286.

358

Sheagren

19. Bovornkitti S, Kangsadal P, Sathirapat P, Oonsombatti P. Reversion and reconversion rate of tuberculin skin reactions in connection with the use of prednisone. Dis Chest 1960; 38:51–55. 20. MacGregor RR, Sheagren JN, Lipsett MB, Wolff SM. Alternate-day prednisone therapy. Evaluation of delayed hypersensitivity responses, control of disease and steroid side effects. N Engl J Med 1969; 280(26):1427–1431. 21. Haanaes OC, Bergmann A. Tuberculosis emerging in patients treated with corticosteroids. Eur J Respir Dis 1983; 64(4):294–297. 22. Cupps TR, Gerrard TL, Falkoff RJ, Whalen G, Fauci AS. Effects of in vitro corticosteroids on B cell activation, proliferation, and differentiation. J Clin Invest 1985; 75(2): 754–761. 23. Settipane GA, Pudupakkam RK, McGowan JH. Corticosteroid effect on immunoglobulins. J Allergy Clin Immunol 1978; 62(3):162–166. 24. Hinshaw L, Peduzzi P, Young E, et al. (and the Veterans Administration Systemic Sepsis Cooperative Study Group). Effect of high-dose glucocorticoid therapy on mortality in patients with clinical signs of systemic sepsis. N Engl J Med 1987; 317:659–665. 25. Bone RC, Fisher CJ Jr., Clemmer TP, Slotman GJ, Metz CA, Balk RA. A controlled clinical trial of high-dose methylprednisolone in the treatment of severe sepsis and septic shock. N Engl J Med 1987; 317(11):653–658. 26. Roujeau JC. Pulse glucocorticoid therapy: the ‘‘big shot’’ revisited. Arch Dermatol 1996; 132(12):1499–1502. 27. Sabir S, Werth VP. Pulse glucocorticoids. Dermatol Clin 2000; 18(3):437–446, viii–ix. 28. Mignogna MD, Lo Muzio L, Ruoppo E, Fedele S, Lo Russo L, Bucci E. High-dose intravenous ‘‘pulse’’ methylprednisone in the treatment of severe oropharyngeal pemphigus: a pilot study. J Oral Pathol Med 2002; 31(6):339–344. 29. Friedli A, Labarthe MP, Engelhardt E, Feldmann R, Salomon D, Saurat JH. Pulse methylprednisolone therapy for severe alopecia areata: an open prospective study of 45 patients. J Am Acad Dermatol 1998; 39(4 Pt 1):597–602; Badsha H, Edwards CJ. Intravenous pulses of methylprednisolone for systemic lupus erythematosus. Semin Arthritis Rheum 2003; 32(6):370–377. 30. Chrousos GA, Kattah JC, Beck RW, Cleary PA. Side effects of glucocorticoid treatment: experience of the Optic Neuritis Treatment Trial. JAMA 1993; 269(16):2110–2112. 31. Fauci AS, Dale DC, Balow JE. Glucocorticosteroid therapy: mechanisms of action and clinical considerations. Ann Intern Med 1976; 84(3):304–315. 32. Schleimer RP, Bochner BS. The effects of glucocorticoids on human eosinophils. J Allergy Clin Immunol 1994; 94(6 Pt 2):1202–1213. 33. Packard BD, Weiler JM. Steroids inhibit activation of the alternative-amplification pathway of complement. Infect Immun 1983; 40(3):1011–1014. 34. Engelman RM, Rousou JA, Flack JE III, Deaton DW, Kalfin R, Das DK. Influence of steroids on complement and cytokine generation after cardiopulmonary bypass. Ann Thorac Surg 1995; 60(3):801–804. 35. McGowan JE Jr, Chesney PJ, Crossley KB, LaForce FM. Guidelines for the use of systemic glucocorticosteroids in the management of selected infections. Working Group on Steroid Use, Antimicrobial Agents Committee, Infectious Diseases Society of America. J Infect Dis 1992; 165(1):1–13. 36. Lebel MH, Freij BJ, Syrogiannopoulos GA, et al. Dexamethasone therapy for bacterial meningitis. Results of two double-blind, placebo-controlled trials. N Engl J Med 1988; 319(15):964–971. 37. de Gans J, van de Beek D, European Dexamethasone in Adulthood Bacterial Meningitis Study Investigators. Dexamethasone in adults with bacterial meningitis. N Engl J Med 2002; 347(20):1549–1556. 38. Chaudhuri A. Adjunctive dexamethasone treatment in acute bacterial meningitis. Lancet Neurol 2004; 3(1):54–62.

Infections in Patients on Steroids in the Critical Care Unit

359

39. Tunkel AR, Hartman BJ, Kaplan SL, et al. Practice guidelines for the management of bacterial meningitis. Clin Infect Dis 2004; 39(9):1267–1284. 40. Thompson BT. Glucocorticoids and acute lung injury. Crit Care Med 2003; 31(suppl 4): S253–S257. 41. Minneci PC, Deans KJ, Banks SM, Eichacker PQ, Natanson C. Meta-analysis: the effect of steroids on survival and shock during sepsis depends on the dose. Ann Intern Med 2004; 141(1):47–56. 42. Luce JM. Physicians should administer low-dose corticosteroids selectively to septic patients until an ongoing trial is completed. Ann Intern Med 2004; 141(1):70–72.

PART III: SPECIAL PROBLEMS IN THE CRITICAL CARE UNIT

18 Fever and Rash in the Critical Care Unit Lee S. Engel, Charles V. Sanders, and Fred A. Lopez Department of Medicine, Louisiana State University Health Science Center, New Orleans, Louisiana, U.S.A.

INTRODUCTION There are numerous potential etiologic agents that can cause the syndrome of fever and rash. Skin manifestations may be an early sign of a life-threatening infection. The ability to rapidly identify the cause of fever and rash in critically ill patients is essential for the proper management of the patient and protection of the healthcare worker(s) providing care for that patient. A rapid method to narrow the potential life-threatening causes of fever and rash has been described by Cunha (1). Patients from the community who are ill enough to be admitted to the critical care unit with fever and rash from outside the hospital will most likely have meningococcemia, Rocky Mountain spotted fever (RMSF), community-acquired toxic shock syndrome (TSS), severe drug reactions, severe bacteremia, Vibrio vulnificus septicemia, gas gangrene, arboviral hemorrhagic fevers, dengue infection, or measles (Table 1). Patients who develop fever and rash after admission to the hospital will most commonly have drug reactions, staphylococcal bacteremia from central lines, exacerbations of systemic lupus erythematosis, or postoperative TSS. The traditional approach to the patient with fever and rash is based on the characteristic appearance of the rash (2,3). The most common types of rash include petechial, maculopapular, vesicular, erythematous, and nodular. Although there can be overlaps in presentation, most causes of fever and rash can be grouped into one specific form of cutaneous eruption (3). A systematic approach requires a thorough history that includes patient age, seasonality, travel, geography, immunizations, childhood illnesses, sick contacts, medications, and the immune status of the host. A detailed history, physical exam, and characterization of the rash will help the clinician reduce the number of possible etiologies. Appropriate laboratory testing will also assist in delineating the cause of fever and rash in the critically ill patient.

361

362

Engel et al.

Table 1 Etiology of Rash and Fever Based on Admission Status Rash and fever on admission to the critical care unit Meningococcemia RMSF Overwhelming pneumococcal or staphylococcal sepsis TSS Epidemic typhus Typhoid fever Measles Arboviral hemorrhagic fever Gas gangrene (Clostridial myenocrosis) Dengue SLE Vibrio vulnificus

Rash and fever after admission to the critical care unit Drug reaction Nosocomial acquired toxic shock syndromes Nosocomial staphylococcal sepsis ‘‘Surgical’’ scarlet fever Cholesterol emboli syndrome

Abbreviations: SLE, systemic lupus erythematosis; RMSF, Rocky Mountain spotted fever; TSS, toxic shock syndrome. Source: Adapted from Ref. 1.

History A comprehensive history of the events leading up to the development of fever and rash will significantly aid in the determination of the etiology of the illness. Several initial questions should be answered before taking a complete history (4,5). 1. Can the patient or someone who is with the patient provide a history? 2. Does the patient require cardiopulmonary resuscitation? 3. Are special isolation precautions needed? For example, patients with meningitis due to Neisseria meningitidis will need droplet precautions, whereas patients with Varicella infections will need airborne and contact precautions (Table 2). Health care workers should exercise universal precautions with all patients. Gloves should be worn during the examination of the skin whenever an infectious etiology is considered. 4. Are the skin lesions suggestive of a disease process that requires immediate antibiotic therapy? Patients with infections suggestive of N. meningitidis, RMSF, bacterial septic shock, TSS, or V. vulnificus will need urgent medical and possibly surgical treatment to improve their chance of survival. 5. Does the patient have an exotic disease due to travel or bioterrorism? Agents, such as smallpox and viral hemorrhagic fevers (i.e., Ebola and Marburg) produce a generalized rash, whereas plague and anthrax may produce localized lesions. Again, isolation precautions will need to be addressed (Table 2). After the preliminary evaluation of the patient, the physician can obtain a more thorough history including history of present illness and previous medical, social, and family histories. Specific questions about the history of the rash itself are often helpful in determining its etiology (Table 3). Such questions should include time of onset, site of onset,

Fever and Rash in the Critical Care Unit

363

Table 2 Transmission-Based Precautions for Hospitalized Patients Standard precautions Use standard precautions for the care of all patients Airborne precautions In addition to standard precautions, use airborne precautions for patients known or suspected to have serious illnesses transmitted by airborne droplet nuclei. Examples of such illnesses include: Measles Varicella (including disseminated zoster)a Tuberculosisb Droplet precautions In addition to standard precautions, use droplet precautions for patients known or suspected to have serious illnesses transmitted by large particle droplets. Examples of such illnesses include: Invasive Haemophilus influenzae type b disease, including meningitis, pneumonia, epiglottitis, and sepsis Invasive Neisseria meningitidis disease, including meningitis, pneumonia, and sepsis Other serious bacterial respiratory infections spread by droplet transmission, including: Diphtheria (pharyngeal) Mycoplasma pneumonia Pertussis Pneumonic plague Streptococcal pharyngitis, pneumonia, or scarlet fever in infants and young children Serious viral infections spread by droplet transmission, including those caused by: Adenovirus Influenza Mumps Parvovirus B19 Rubella Contact precautions In addition to standard precautions, use contact precautions for patients known or suspected to have serious illnesses easily transmitted by direct patient contact or by contact with items in the patient’s environment. Examples of such illnesses include: Gastrointestinal, respiratory, skin, or wound infections or colonization with multidrug-resistant bacteria judged by the infection control program, based on current state, regional, or national recommendations, to be of special clinical and epidemiologic significance Enteric infections with a low infectious dose or prolonged environmental survival, including those caused by: Clostridium difficile For diapered or incontinent patients: enterohemorrhagic E. coli 0157:H7, Shigella, hepatitis A, or rotavirus Respiratory syncytial virus, parainfluenza virus, or enteroviral infections in infants, and young children Skin infections that are highly contagious or that may occur on dry skin, including: Diphtheria (cutaneous) Herpes simplex virus (neonatal or mucocutaneous) Impetigo Major (non-contained) abscesses, cellulitis, or decubiti Pediculosis Scabies (Continued)

364

Table 2

Engel et al.

Transmission-Based Precautions for Hospitalized Patients (Continued )

Staphylococcal furunculosis in infants and young children Zoster (disseminated or in the immunocompromised host) Viral/hemorrhagic conjunctivitis Viral hemorrhagic infections (Ebola, Lassa, or Marburg) Note: CDC infection control guidelines reprinted from Garner JS and the Hospital Infection Control Practices Advisory Committee. a Certain infections require more than one type of precaution. b See Centers for Disease Control and Prevention. Source: From Refs. 6 and 7.

change in lesions, symptoms associated with the rash (i.e., itching, burning, numbness, and tingling), provoking factors, previous rashes, and prior treatments. The physical exam should focus on the patient’s vital signs, general appearance, and the assessment of lymphadenopathy, nuchal rigidity, neurological dysfunction, hepatomegaly, splenomegaly, arthritis, and mucosal membrane lesions (Table 4) (3,4). Skin examination to determine type of the rash (Table 5) includes evaluation of distribution pattern, arrangement, and configuration of lesions. The remainder of this chapter will provide a diagnostic approach to patients with fever and rash based on the characteristics of the rash. Several clinically relevant causes of each type of rash associated with fever are described in brief detail. Table 3 Fever and Rash: History Age of patient Season of the year Type of prodrome associated with current illness History of drug or antibiotic allergies Medications taken with in the past 30 days (prescription or nonprescription) Drug ingestion Exposure to febrile or ill persons within the recent past Prior illness Occupational exposures Sun exposures Recent travel Exposure to wild or rural habitats Exposure to insects, arthropods, or wild animals Exposure to pets Immunizations Exposure to sexually transmitted diseases HIV risk factors (intravenous drug use, unprotected sex, sexual orientation) Site of rash onset Factors effecting immunological status (chemotherapy, steroid use, hematological malignancy, solid organ or bone marrow transplant, asplenia) Valvular heart disease Rate of rash development (slow vs. fast) Direction of rash spread (centrifugal vs. centripetal) Evolution of rash (Has the rash changed?) Relationship between rash and fever Presence or absence of pruritus Previous treatment of the rash (topical or oral therapies) Source: Adapted from Refs. 5 and 8.

Fever and Rash in the Critical Care Unit

365

Table 4 Fever and Rash: Physical Examination Vital signs Temperature Pulse Respiration Blood pressure General appearance Alert Acutely ill Chronically ill Signs of toxicity Adenopathy/location of adenopathy Presence of mucosal, conjunctival, or genital lesions hepatosplenomegaly Arthritis Nuchal rigidity/Neurological dysfunction Features of rash Type of primary rash lesion (Table 5) Presence of secondary lesions Presence of desquamation Presence of excoriations Configuration of individual lesions Arrangement of lesions Distribution pattern: exposed areas; centripetal vs. centrifugal Source: Adapted from Refs. 5 and 8.

PETECHIAL AND PURPURIC RASHES Petechiae are produced by extravasation of red blood cells and are less then 3 mm in diameter. Petechiae appear as small red or brown spots on the skin. Purpura or ecchymoses are lesions that are larger than 3 mm and often form when petechiae coalesce. Neither petechial nor purpuric lesions blanch when pressure is applied. Infections associated with diffuse petechiae are generally among the most life threatening and require urgent evaluation and management. There are many infectious causes of these lesions (Table 6); several of the most dangerous include meningococcemia, rickettsial infection, and bacteremia (1,3,8).

Table 5 Type of Rash Lesions Macule Papule Plaque Nodule Pustule Vesicle Bulla

Circumscribed flat lesion that differs from surrounding skin by color. Patches are very large macular lesions Circumscribed, solid, elevated skin lesion that is palpable and smaller then 0.5 cm in diameter Large, solid, elevated skin lesion that is palpable and greater the 0.5 cm in diameter, often formed by confluence of papules Circumscribed, solid, palpable skin lesion with depth as well as elevation Circumscribed raised lesion containing pus Circumscribed elevated, fluid filled lesion less then 0.5 cm in diameter Circumscribed elevated, fluid filled lesion greater then 0.5 cm in diameter

Source: Adapted from Refs. 5 and 9.

366

Engel et al.

Table 6 Etiology of Rash and Fever Based on Type of Rash Purpura or Petechiae Meningococcemia RMSF Gonococcemia Staphylococcal/pneumococcal sepsis Pseudomonal sepsis Bacterial endocarditis Typhus Allergic vasculitis Echovirus 9 Measles Centrally Distributed Maculopapular Rash Viral exanthems (rubeola, rubella, erythema infectiousum, roseola) Lyme disease Drug reactions Peripherally Distributed Maculopapular Rash Erythema multiforme (Table 7) Secondary syphilis Diffuse erythema with desquamation Scarlet fever TSS Scalded skin syndrome Kawasaki disease Erlichiosis TEN Streptococcus viridans bacteremia Vesicular, Bullous, or Pustular Rash Varicella Herpes zoster Herpes simplex Staphylococcus aureus bacteremia Vibrio vulnificus Rickettsia akari Nodular Rash Erythema Nodosum (Table 8) Disseminated fungal infections (Candida, Cryptococcus, Blastomycosis, Histoplasma, Coccidiodes, and Sporothrix) Nocardia Mycobacteria Abbreviations: TEN, toxic epidermal necrolysin; RMSF, Rocky Mountain spotted fever; TSS, toxic shock syndrome. Source: Adapted from Refs. 1, 3, 5, and 8.

Acute Meningococcemia N. meningitidis is the leading cause of bacterial meningitis in children and young adults (10). Bacterial meningitis associated with a petechial or purpuric rash should always suggest meningococcemia (1). The diagnosis of meningococcemia is more difficult to make when meningitis is not present. Meningococcemia can occur sporadically or in epidemics, and is more commonly diagnosed during the winter months. The risk of infection is highest in infants,

Fever and Rash in the Critical Care Unit

367

asplenic patients, alcoholics, patients with complement deficiency, and persons who live in dormitories (coeds, military personnel, or prisoners). Initial symptoms include cough, headache, sore throat, nausea, and vomiting. Acute meningococcemia progresses rapidly and patients typically appear ill with high spiking fevers, tachypnea, tachycardia, mild hypotension, and a characteristic petechial rash (11,12). Signs and symptoms of meningeal irritation, such as headache, vomiting, and change in conscious state occur in up to 88% of patients with meningococcemia (11,13). The rash associated with meningococcemia begins within 24 hours of clinical illness. The petechia enlarges rapidly, becoming papular and then purpuric. Lesions most commonly occur on the extremities and trunk but may also be found on the head and mucous membranes (5). The development of lesions on the palms and soles is usually a late finding (1). Purpuric skin lesions have been described in 60% to 100% of meningococcemia cases and are most commonly seen at presentation (Fig. 1) (14,15). Histological studies demonstrate diffuse vascular damage, fibrin thrombi, vascular necrosis, and perivascular hemorrhage in the involved skin and organs. The skin lesions associated with meningococcal septic shock are thought to result from an acquired or transient deficiency of protein C and/or protein S (16). Meningococci are present in endothelial cells and neutrophils, and smears of skin lesions are positive for gram-negative diplococci in many cases (17,18). The diagnosis of meningococcemia is also aided by culturing the petechial lesions. Blood cultures shouldbe drawn. Admission laboratorydata usually demonstrate a leukocytosis and thrombocytopenia. Patients with meningococcemia but without meningitis will have a normal cerebrospinal fluid (CSF) profile. If meningococcal meningitis is present, the CSF culture is usually positive although the gram-stain may be negative. Typically, the CSF-associated glucose is low and the protein elevated. Chronic Meningococcemia Chronic meningococcemia is rare, and its lesions differ from those seen in acute meningococcemia. Patients present with intermittent fever, rash, arthritis, and arthralgias occurring over a period of several weeks to months (19,20). The lesions of chronic

Figure 1 Purpuric skin lesions on an infant with meningococcal septicemia. Source: Courtesy of the CDC.

368

Engel et al.

meningococcemia are usually pale to pink colored macules and/or papules typically located around a painful joint or pressure point. Nodules may develop in the lower extremities. The lesions of chronic meningococcemia develop during periods of fever and fade when the fevers dissipate. These lesions (in contrast to those of acute meningococcemia) rarely demonstrate the bacteria on Gram stain or histology (5,8). Rocky Mountain Spotted Fever RMSF, the most lethal rickettsial disease in the United States, is caused by Rickettsia rickettsii (21–24). Infection occurs after a bite by the tick vector, Dermacentor. RMSF is more common in men, occurs most often between April and September, and is most prevalent in Oklahoma and the southern Atlantic states. The onset of RMSF can be abrupt with fever, headache, myalgias, shaking chills, photophobia, and nausea. Patients may have periorbital edema, conjunctival suffusion, and localized edema involving the dorsum of the hands and feet (1). A notable clinical finding is a pulse-temperature deficit (i.e., relative bradycardia during fever). Localized abdominal pain secondary to liver involvement, renal failure manifested by acute tubular necrosis, pancreatitis, left ventricular failure, adult respiratory distress syndrome (ARDS), and mental confusion or deafness may also be noted (1). The rash usually begins about four to five days after the start of the illness. The lesions are initially maculopapular and evolve into petechiae within two to four days. Characteristically, the rash starts on the wrists, forearms, ankles, palms, and soles and then spreads centripetally to involve the arms, thighs, trunk, and face (Fig. 2). Centripetal evolution of the rash occurs 6 to 18 hours after the rash develops. Prompt treatment with tetracycline decreases mortality (25,26). Routine admission tests may demonstrate a normal or decreased peripheral white blood cell (WBC) count and thrombocytopenia. The total bilirubin and serum transaminases may be elevated. If pancreatitis is present, the serum amylase will be

Figure 2 Childs right hand and wrist demonstrating the characteristic spotted rash of Rocky Mountain spotted fever. Source: Courtesy of the CDC Public Health Image Library.

Fever and Rash in the Critical Care Unit

369

elevated. Patients who develop renal failure may demonstrate a rise in blood urea nitrogen (BUN) and creatinine suggestive of prerenal azotemia secondary to intravascular volume deficit. When the central nervous system is involved, the CSF profile will demonstrate a mild pleocytosis, normal glucose and protein concentrations, and negative Gram stain and culture. Blood cultures will also be negative in RMSF. Serological studies will demonstrate the presence of antibodies seven to 10 days after symptoms start. Septic Shock The yearly incidence of sepsis has been increasing about 9% a year and accounts for 2% of all hospital admissions (27). The peak incidence of septic shock occurs in patients who are in their seventh decade of life (28). Risk factors for sepsis include cancer, immunodeficiency, chronic organ failure, and iatrogenic factors. Sepsis develops from infections of the chest, abdomen, genitourinary system, and primary bloodstream in more than 80% of cases (28,29). Symmetric peripheral gangrene or purpura fulminans is a cutaneous syndrome associated with septic shock secondary to N. meningitidis or Streptococcus pneumoniae. This syndrome is usually preceded by petechiae, ecchymosis, purpura, and acrocyanosis. Acrocycanosis, another cutaneous manifestation of septic shock, is a grayish color of the skin, which occurs on the lips, legs, nose, ear lobes, and genitalia, and does not blanch on pressure. Bacteria are usually absent in smears obtained from these skin lesions. Sepsis is defined as systemic inflammatory response syndrome with documented infection. Patients with sepsis will therefore have a documented site of infection and display two or more of the following: body temperature > 38.5 C or < 35 C; heart rate >90 beats/min; respiratory rate >20 breaths/min; arterial CO2 tension 12,000/mm3 or WBC 10%. With severe sepsis, patients begin to demonstrate areas of mottled skin, capillary refill time more than three seconds, decreased urine output, changes in mental status, thrombocytopenia, disseminated intravascular coagulation (DIC), cardiac dysfunction, and ARDS. When patients can no longer maintain a systemic mean arterial blood pressure of 60 mmHg or require a vasopresser agent, then they are said to be in septic shock. Mortality varies from 35% to 70% depending on patients’ age, sex, ethnic origin, co-morbidities, and presence of acute lung injury or ARDS, as well as on whether the infection is nosocomial or polymicrobial, or whether the causative agent is a fungus (28,29). Gram-negative infections are responsible for 25% to 30% of cases of septic shock, whereas gram-positive infections now account for 30% to 50% of the cases of septic shock. Multidrug-resistant bacteria and fungi are increasingly reported as causes of sepsis (28,29). The diagnosis of septic shock requires a causal link between infection and organ failure (28). Some patients may have clinically obvious infection, such as purpura fulminans, cellulitis, TSS, pneumonia, or a purulent wound. Without an obvious source of infection, diagnosis will require the recovery of pathogens from blood or tissue cultures. Unfortunately, cultures are negative in 30% of these cases. Bacterial Endocarditis Infective endocarditis is classified as acute or subacute based on the tempo and severity of the clinical presentation (30). The characteristic lesion is a vegetation composed of platelets, fibrin, microorganisms, and inflammatory cells on the heart

370

Engel et al.

Figure 3 Subungual hemorrhages in an adult patient with Group B streptococcal endocarditis. Source: Courtesy of Lee S. Engel.

valve. Conditions associated with endocarditis include injection drug use, poor dental hygiene, long-term hemodialysis, diabetes mellitus, HIV infection, long-term indwelling venous catheters, mitral valve prolapse with regurgitation, rheumatic heart disease, other underlying valvular diseases, and prosthetic valves (31–33). Organisms associated with endocarditis include Staphylococcus aureus, viridans streptococci, enterococci, gram-negative bacilli (including the HACEK organisms), and fungi. Nonspecific symptoms and signs of endocarditis include fever, arthralgias, wasting, unexplained heart failure, new heart murmurs, pericarditis, septic pulmonary emboli, strokes, and renal failure (34). Skin lesions occur less frequently today than they once did but aid in the diagnosis if present (34). Cutaneous manifestations of endocarditis include splinter hemorrhages (Fig. 3), petechiae, Osler’s nodes, and Janeway lesions. Petechiae are the most common skin lesions seen during endocarditis. The petechiae are small, flat, and reddish brown and do not blanch with pressure. They frequently occur in small crops and are usually transient. They are often found on the heels, shoulders, legs, oral mucous membranes, and conjunctiva. Osler’s nodes may be seen in patients with subacute bacterial endocarditis. These nodules are tender, indurated, and erythematous. They occur most commonly on the pads of the fingers and toes, are transient, and resolve without the development of necrosis. The histology of these lesions demonstrates microabscesses and microemboli. Janeway lesions are small, painless, erythematous macules that are found on the palms and soles. These lesions can be seen with both acute and subacute endocarditis. Histological analysis reveals microabscesses with neutrophil infiltration. Disseminated Gonococcal Infection Disseminated gonococcal infections (DGI) result from gonococcal bacteremia and occur in 1% to 3% of patients with untreated Neisseria gonorrhea–associated

Fever and Rash in the Critical Care Unit

371

Figure 4 Cutaneous lesions on the left ankle and calf of a patient with disseminated Neisseria gonorrhea infection. Source: Courtesy of the CDC/Dr. S.E. Thompson, Public Health Image Library.

mucosal infection (35–37). DGI is most often seen in young women during menses or pregnancy (38). Most patients will present with fever, rash, polyarthritis, and tenosynovitis (36). Skin lesions, which occur in 50% to 70% of patients with DGI, are the most common manifestation (38). The rash usually begins on the first day of symptoms and becomes more prominent with the onset of each new febrile episode (39). The lesions begin as tiny red papules or petechiae (1–5 mm in diameter) that evolve to a vesicular and then pustular form (Fig. 4). The pustular lesions develop a gray, necrotic center with a hemorrhagic base (36,39). The rash of DGI tends to be sparse and widely distributed, and the distal extremities are most commonly involved. Gram stain of the skin lesions rarely demonstrates organisms. Clinical clues of DGI include the symptoms of fever, rash, and arthritis/tenosynovitis. Early in the infection, blood cultures may be positive; later, synovial joint fluid from associated effusions may yield positive cultures. Smears of the cervix and urethral exudates may also yield positive results. Capnocytophaga Infection Capnocytophaga canimorsus is a fastidious gram-negative bacteria that is part of the normal gingival flora of dogs and cats (40,41). Human infections are associated with dog or cat bites, cat scratches, and contact with wild animals (40,41). Predisposing

372

Engel et al.

factors include trauma, alcohol abuse, steroid therapy, chronic lung disease, and asplenia (40,41). The clinical syndrome consists of fever, DIC, necrosis of the kidneys and adrenal glands, thrombocytopenia, hypotension, and renal failure. The mortality rate approaches 25%. Skin lesions occur in 50% of infected patients often progressing from petechiae to purpura to cutaneous gangrene (42). Other dermatologic lesions include macules, papules, painful erythema, or eschars. Clinical clues include a compatible clinical syndrome and a history of a dog- or cat-inflicted wound. Diagnosis depends on the culture of the bacteria from blood, tissues, or other body fluids. More prompt diagnosis may be made by Gram staining the buffy coat. Dengue Dengue is a flavivirus comprised of four serotypes, i.e., DEN-1, DEN-2, DEN-3, and DEN-4. Dengue viruses are transmitted from person to person through infected female Aedes mosquitoes. The mosquito acquires the virus by taking a blood meal from an infected human or monkey. The virus incubates in the mosquito for 7 to 10 days before it can transmit the infection. More then 2.5 billion people are at risk for dengue infections worldwide (43). Dengue fever (also known as breakbone fever or dandy fever) is a short-duration, nonfatal disease characterized by the sudden onset of headache, retro-orbital pain, high fever, joint pain, and rash (43,44). The rash manifests as palpable pinpoint petechiae that begin centrally and spread peripherally (1). Dengue fever lasts about seven days. Recovery from infection provides lifelong immunity to that serotype but does not preclude patients from being infected with the other serotypes of dengue virus, i.e., secondary infections. Dengue hemorrhagic fever and dengue TSS are two deadly complications of dengue viral infection that occur during secondary infection. Dengue hemorrhagic fever is characterized by hemorrhage, thrombocytopenia, and plasma leakage. Dengue shock syndrome includes the additional complications of circulatory failure and hypotension (43,44). The incubation period for dengue virus infections is 3 to 14 days. If a patient presents greater than two weeks after visiting an endemic area, dengue is much less likely (45). Laboratory abnormalities include neutropenia followed by lymphocytosis, hemoconcentration, thrombocytopenia, and an elevated aspartate aminotransferase in the serum (46). The diagnosis of dengue virus–associated infection can be accomplished by polymerase chain reaction (PCR), detection of antidengue virus immunoglobulin M (IgM), centrifugation amplification to enhance virus isolation, or flow cytometry for early detection of cultured virus (47). MACULOPAPULAR RASH Lyme Disease Lyme disease is the most common tick vector–associated disease in the United States (48–50). Lyme disease is caused by the spirochete Borrelia burgdorferi, which is transmitted by the tick Ixodes. Lyme disease is endemic in the northeastern, midAtlantic, north central, and far western regions of the United States. The disease has a bimodal age distribution, with peaks in patients younger than 15 and older than 29 years of age (51). Most infections occur between May and September.

Fever and Rash in the Critical Care Unit

373

Lyme disease has three stages: early localized, early disseminated, and late disease. Early localized disease is characterized by erythema migrans (EM), which forms 7 to 10 days following the tick bite (52). EM occurs in 60% to 80% of the cases and begins as a small red papule at the site of the bite. The lesion expands centrifugally and can get as large as 70 cm in diameter. The lesion develops central clearing in 30% of cases (Fig. 5). If untreated, the lesions resolve over several weeks. Other symptoms associated with early-localized disease include fatigue, myalgias, arthralgias, headache, fever, and chills. Early disseminated disease occurs days to weeks after the tick bite. Patients may not recall having had the typical EM rash. Patients at this stage can present with lymphocytic meningitis, cranial nerve palsies, mild pericarditis, atrial-ventricular block, arthritis, generalized or regional adenopathy, conjunctivitis, iritis, hepatitis, and painful radiculoneuritis followed by decreased sensation, weakness, and absent reflexes (48,49,53). Disseminated skin lesions, when present, are similar to EM but smaller and usually multiple in number. Late disease is characterized by chronic asymmetric oligoarticular arthritis that involves the large joints (most often the knee). The central nervous system may also be affected, manifesting as subacute encephalopathy, axonal polyneuropathy, or leukoencephalopathy. Diagnosis is based on the history and physical exam. Serology is confirmatory but takes four to six weeks after the onset of symptoms to become positive. CSF should be obtained if neurological signs are present. Synovial fluid can be evaluated if arthritis is present. Drug Reactions Drugs cause adverse skin reactions in 2% to 3% of hospitalized patients (54). Classic drug reactions include urticaria, angioedema, exanthems, vasculitis, exfoliative

Figure 5 Characteristic rash, erythema migrans, on the arm of a patient with Lyme disease. Source: Courtesy of Allen C. Steere, www.forstryimages.org and National Institute of Health, U.S. National Library of Medicine.

374

Engel et al.

dermatitis/erythroderma, erythema multiforme, Stevens–Johnson syndrome, and toxic epidermal necrolysis (TEN) (54,55). There is no predilection for age, gender, or race (8). Diagnosis of a drug reaction is based on a patient’s previous reaction to the drug, ruling out alternate etiological causes of the rash, timing of events, drug levels, or evidence of overdose, patient reaction to drug discontinuation, and patient reaction to re-challenge. Drug Exanthems Exanthems are the most common skin reaction to drugs. The rash usually appears within the first two weeks after the offending drug is started and resolves within days after the drug is stopped. The rash is often described as morbilliform, macular, and/ or papular eruption. Pruritus is the most common associated symptom of druginduced rash. Low-grade fever and peripheral blood eosinophilia may also occur in association with drug exanthems. Erythema Multiforme Erythema multiforme is an acute, self-limited, peripheral eruptive maculopapular rash that is characterized by a target lesion. Erythema multiforme most often affects persons 20 and 30 years of age and has a predilection for men. The rash begins as a dull-red macular eruption that evolves into papules and the characteristic target lesion. Target lesions are often found on the palms, soles, knees, and elbows. Vesicles and bullae occasionally develop in the center of the papules (8). There are many causes of this disorder (Table 7). Erythema multiforme is classified as minor or major (8). Bullae and systemic symptoms are absent in erythema multiforme minor. The rash rarely affects the mucous membranes and is usually limited to the extensor surfaces of the extremities. The most common cause of erythema multiforme minor is herpes simplex virus (HSV). Erythema multiforme major is usually caused by drug reactions. Mucous membranes are involved, and the eruptions often become bullous. Fever, cheilosis, stomatitis, balanitis, vulvitis, and conjunctivitis can also occur (54). Stevens–Johnson Syndrome Stevens–Johnson syndrome is a blistering disorder that is usually more severe than erythema multiforme (57,58). The causes of Stevens–Johnson syndrome are similar to the etiologies of erythema multiforme (Table 7). Patients with Stevens–Johnson syndrome often present with pharyngitis, malaise, and fever. The syndrome evolves over a few days with the evolution of mucous membrane erosions. Small blisters develop on purpuric or atypical target lesions. The blisters eventually result in skin detachment. Stevens–Johnson syndrome affects less than 10% of the total body surface (54,58). Toxic Epidermal Necrolysis TEN is the most serious cutaneous drug reaction and is defined by blistering of over 30% of the total body surface area. More than one mucous membrane is involved. TEN is usually caused by the same drugs that cause erythema multiforme (Table 7), and its onset is acute. A fever greater than 39 C is often present. Intestinal and pulmonary involvement predicts a poor outcome (54,55). The diagnosis of Stevens–Johnson syndrome and TEN is made by skin biopsy. Sections of frozen skin will demonstrate full-thickness epidermal necrosis. Because extensive skin detachment results in massive transepidermal fluid losses, patients with these maladies are managed similarly to patients who have had extensive burn

Fever and Rash in the Critical Care Unit

375

Table 7 Causes of Erythema Multiforme Viral Infections Herpes simplex 1 and 2 Epstein–Barr virus Hepatitis A, B, C Varicella zoster Parvovirus B19 Bacterial Infections Hemolytic streptococci Pneumococcus Staphylococcus species Proteus species Salmonella species Mycobacterium tuberculosis Mycobacterium avium complex Francisella tularensis Vibrio parahaemolyticus Yersinia species Mycoplasma pneumonia Fungal Infections Histoplasma capsulatum Coccidiomycosis Parasitic Infections Trichomonas species Toxoplasma gondii Antibiotics Penicillin Tetracyclines Erythromycin Sulfa drugs Vancomycin Anticonvulsants Barbiturates Carbamazepine Phenytoin Antituberculoids Rafampin Isoniazid Pyrizinamide Other Drugs Allopurinol Fluconazole hydralazine NSAIDs Estrogen Physical factors/contact Sunlight Cold X-ray therapy Tattooing Poison ivy Other factors (Continued)

376 Table 7

Engel et al. Causes of Erythema Multiforme (Continued )

Pregnancy Multiple myeloma Leukemia Collagen diseases Idiopathic (50%) Abbreviations: NSAIDs, nonsteroidal antiinflammatory drugs. Source: Adapted from Ref. 56.

injuries. Sepsis can occur secondary to microbial colonization of denuded skin. Mortality rates are 5% for Stevens–Johnson syndrome and 50% for TEN (54). Secondary Syphilis Syphilis is a systemic disease caused by Treponema pallidum. Syphilis is classified into primary, secondary, early latent, late latent, and tertiary stages. The lesion of primary syphilis, the chancre, usually develops about 21 days after infection and resolves in one to two months. Patients with secondary syphilis can present with rash, mucosal lesions, lympadenopathy, and fever. The rash of secondary syphilis may be maculopapular, papulosquamous, or pustular, and is characteristically found on the palms and the soles (Fig. 6). The diagnosis of syphilis is based on nontreponemal tests (e.g., Venereal Disease Research Laboratory and Rapid Plasma Reagin) and specific treponemal tests

Figure 6 Papulosquamous rash on wrist and hands of patient with secondary syphilis. Source: Courtesy of the CDC/Susan Lindsley, Public Health Image Library.

Fever and Rash in the Critical Care Unit

377

(e.g., fluorescent treponemal antibody absorbed and T. pallidum particle agglutination). The nontreponemal tests are used to screen for disease and follow treatment. The specific treponemal tests are used to rule in the diagnosis of syphilis because false-positive nontreponemal tests can occur. Darkfield examination of skin or mucous membrane lesions can be done to diagnose syphilis definitively during the early stages as well. West Nile Virus West Nile virus (WNV) is transmitted to humans from the bite of an infected mosquito (59). The virus normally circulates between mosquitoes and birds. The first reported outbreak was in New York in 1999, and since then WNV has spread southward and westward (60–63). WNV has become seasonally endemic, with peak activity for transmission from July to October in temperate zones and from April to December in warmer climates (61,63). Though most commonly spread by infected mosquitoes, WNV may also be transmitted by organ transplantation, blood transfusion, and breast milk (64–66). Transplacental infection from mother to fetus has also occurred (64). WNV replicates at the site of inoculation and then spreads to the lymph nodes and bloodstream (67). The majority of human infections, i.e., 80%, are asymptomatic (68). Most patients with symptoms have self-limited West Nile fever. West Nile fever is characterized by acute onset of fever, headache, fatigue, malaise, muscle pain, difficulty concentrating, and neck pain.(69,70). Approximately 57% of patients with West Nile fever will have a transient macular rash on the trunk of the body (69). Neuroinvasive disease develops in less than 1% of infected patients (68). The clinical severity of WNV encephalitis ranges from disorientation to coma to death (71,72). Advanced age is the most significant risk factor for severe neurologic disease. Risk increases 10 times for persons 50 to 59 years of age and 43 times for persons greater than 80 years of age (61,65). Neuroinvasive disease can present as meningitis, encephalitis, or paralysis (68,70,72,73). Patients with WNV encephalitis or focal neurologic findings will often have persistent deficits for months to years (61,72). Advanced age is the most important risk factor for death. The overall case fatality rate for neuroinvasive WNV disease is 9% (61). Diagnosis of WNV disease can be made by a high index of clinical suspicion and detection of WNV-specific immunoglobulin M (IgM) in serum or CSF. The serum IgM can persist for up to eight months; therefore, nucleic amplification tests for WNV, such as reverse transcriptase PCR and real-time PCR may be required to prove that the infection is acute (70,74). Neuroinvasive WNV can be diagnosed by the presence of IgM-specific antibody in the CSF. Patients who have been recently vaccinated for yellow fever or Japanese encephalitis or persons recently infected with the St. Louis encephalitis virus or Dengue virus may have false-positive results on IgM antibody tests for WNV (75).

DIFFUSE ERYTHEMATOUS RASHES WITH DESQUAMATION Toxic Shock Syndrome TSS is characterized by sudden onset of fever, chills, vomiting, diarrhea, muscle aches, and rash. TSS can rapidly progress to severe hypotension and multiorgan dysfunction. The overall case fatality rate is 5%.

378

Engel et al.

The microbial etiology of TSS is usually S. aureus; however, coagulase-negative staphylococci, group A streptococci, and group B streptococci can also produce this syndrome (76–78). TSS is most commonly seen in menstruating women, women using barrier contraceptive devices, persons who have undergone nasal surgery, and patients with postoperative staphylococcal wound infections (79). Initially, cases associated with menstruation accounted for as many as 91% of the total cases (79). Currently, only half of the reported TSS cases are menstrual (80). Staphylococcal TSS Staphylococcal TSS is caused by infection or colonization with toxin-producing bacteria. The most common toxins associated with TSS include toxin 1 and enterotoxin B (81–84). Other toxins that may be involved include enterotoxins A, C, D, E, and H (85). The clinical presentation of TSS was defined by the Centers for Disease Control and Prevention (CDC) in 1981 (4). All patients with TSS have high fever ( >39 C), hypotension, and skin manifestations. Patients may also present with headache, vomiting, diarrhea, myalgias, pharyngitis, conjunctivitis, vaginitis, arthralgias, abdominal pain, or encephalopathy (86–89). The syndrome can progress to shock, DIC, ARDS, and renal failure. The rash of TSS may start as erythroderma that involves both the skin and mucous membranes. The rash is diffuse, red, and macular and may resemble sunburn. The rash can involve the palms and soles. The erythema may be more intense around a surgical wound site. Mucosal involvement can involve the conjunctiva, oropharynx, or vagina (90). One to three weeks after the onset of TSS, the palms and soles may desquamate (Fig. 7) (91). TSS can be divided into menstrual versus nonmenstrual. The majority of menstrual cases of TSS are associated with tampon use (92). Nonmenstrual cases are caused by abscesses, cellulitis, bursitis, postpartum infections, vaginal infections, sinusitis, burn wounds, insect bites, and surgical procedures (88,93). The diagnosis of TSS is based on the CDC Diagnostic criteria (4). Although S. aureus is isolated from mucosal or wound sites in 80% to 90% of patients with TSS,

Figure 7 Desquamation of left palm of a patient with toxic shock syndrome. Source: Courtesy of the CDC, Public Health Image Library.

Fever and Rash in the Critical Care Unit

379

this criterion is not required for diagnosis. S. aureus is only recovered from blood cultures 5% of the time (92). Other laboratory abnormalities may include hypocalcemia, elevated liver enzymes, elevated creatinine, thrombocytopenia, pyuria, and proteinuria (90). Streptococcal Toxic Shock Syndrome The clinical picture of TSS caused by Group A and B streptococcus is similar to that caused by S. aureus. Skin and soft-tissue infections are often the source of invasive Group A and B streptococci (76,78). Minor trauma, injuries resulting in hematoma or bruising, surgery, viral infections, and use of nonsteroidal anti-inflammatory drugs are associated with the development of severe streptococcal infections (78). One particular difference from staphylococcal-associated TSS is that streptococci can frequently (60% of the time) be isolated from blood culture (94). The mortality rates for streptococcal TSS are five times higher than those for the staphylococcal TSS (95). Staphylococcal Scalded Skin Syndrome Staphylococcal scalded skin syndrome (SSSS) describes a spectrum of superficial blistering skin disorders caused by S. aureus strains that produce exfoliative toxins (96). The clinical spectrum of SSSS includes a localized form, bullous impetigo, and a generalized form, pemphigous neonatorum. The exfoliative toxins are also known as epidermolytic toxins, epidermolysins, and exfoliatins. Production of exfoliative toxin occurs in 5% of all S. aureus strains (97,98). The two main exfoliative toxins are ETA and ETB (99–101). More recently, two new toxins, ETC and ETD, have been identified (101). Bullous impetigo (also known as bullous varicella or measles pemphigoid) presents with a few localized, fragile, superficial blisters that are filled with colorless, purulent fluid (102). The lesions re-epithelialize in five to seven days. This form of SSSS is usually only seen in children. Typically, there are no associated systemic symptoms. The lesions are located in the area of the umbilicus and perineum in infants and over the extremities in older children (103). The generalized form of SSSS is termed pemphigus neonatorum or Ritter’s disease. Risk factors for development in adults include renal dysfunction, lymphoma, and immunosuppression (96,103,104). Patients with pemphigus neonatorum present with fever, erythema, malaise, and irritability. Patients then develop large superficial blisters that rupture easily due to friction (96). A positive Nikolsky sign refers to dislodgement of the superficial epidermis when gently rubbing the skin (105). If untreated, the epidermis will slough off leaving extensive areas of denuded skin that are painful and susceptible to infection. Mucous membranes are not affected in SSSS. The mortality rate in children remains below 5%. Potentially fatal complications in infants and young children occur because of the loss of protective epidermis. Hypothermia, dehydration, and secondary infections are the leading causes of morbidity and mortality for these age groups affected by generalized SSSS (106). The mortality for adults with generalized SSSS is 60%, probably due to the associated comorbidities, such as renal dysfunction, immunosuppression, or malignancy found in this population (107). Diagnosis of both generalized and localized SSSS is based on clinical characteristics. A thorough examination looking for foci of infection (pneumonia, abscess, arthritis, endocarditis, sinusitis, etc.) should be undertaken. Unfortunately, in most

380

Engel et al.

cases, no focus is ever found (96). Blood cultures are usually negative because toxins are produced at a distant site (103,108). A number of different tests, including PCR, enzyme-linked immunosorbent assays, radioimmune assays, and reverse latex agglutination assays, can be used to demonstrate toxin production by S. aureus (109). The diagnostic challenge is that bacteria must first be isolated. When the diagnosis is uncertain, a skin biopsy may be the optimal test. The biopsy typically reveals mid-epidermal splitting at the level of the zona granulosa without cytolysis, necrosis, or inflammation (110). Staphylococci may be seen in bullous lesions of localized disease but rarely in the bullous lesions of generalized disease (104). Scarlet Fever Scarlet fever is the result of infection with a Streptococcus pyogenes strain (i.e., Group A streptococcus) that produces a pyrogenic exotoxin (ethrogenic toxin). There are three different toxins, types A, B, and C, which are produced by 90% of these strains. Scarlet fever follows an acute infection of the pharynx/tonsils or skin (8). It is most common in children between the ages of 1 and 10 years (95). The rash of scarlet fever starts on the head and neck, followed by expansion to the trunk and then extremities (8,111). The rash is erythematous and diffuse, and blanches with pressure. There are numerous papular areas in the rash that produce a sandpaper-type quality. On the antecubital fossa and axillary folds, the rash has a linear petechial character referred to as Pastia’s lines (111). The rash varies in intensity but usually fades in four to five days. Diffuse desquamation occurs after the rash fades (111). Diagnosis of scarlet fever can usually be made on a clinical basis. Confirmation of the diagnosis is supported by isolation of Group A streptococci from the pharynx and serologies (95). Kawasaki Disease Kawasaki disease (KD) is an acute, self-limited, systemic vasculitis of childhood (112–114). KD was first described by Tomisaku Kawasaki in Japan in 1961 (112) and is the predominant cause of pediatric-acquired heart disease in developed countries (114). The signs and symptoms evolve over the first 10 days of illness and then gradually resolve spontaneously in most children. The diagnostic criteria for classical KD include the following (112): 1. Fever for five days or more that does not remit with antibiotics and is often resistant to antipyretics. 2. Presence of at least four of the following conditions: a. Bilateral (nonpurulent) conjunctivitis. b. Polymorphous rash. c. Changes in the lips and mouth: reddened, dry, or cracked lips; strawberry tongue; diffuse erythema of oral or pharyngeal mucosa. d. Changes in the extremities: erythema of the palms or soles; indurative edema of the hands or feet; desquamation of the skin of the hands, feet, and perineum during convalescence. e. Cervical lymphadenopathy: lymph nodes more than 15 mm in diameter. 3. Exclusion of disease with a similar presentation, such as scalded skin syndrome, TSS, viral exanthems, etc.

Fever and Rash in the Critical Care Unit

381

Other clinical features include intense irritability ( possibly due to cerebral vasculitis), sterile pyuria, and upper respiratory symptoms (114). The major morbidity of KD is the development of coronary artery aneurysm(s), which occur in 25% of the cases. There are no specific or sensitive tests that can be used to diagnose KD. The diagnosis is made by clinical assessment of the above criteria. The cause of KD is unknown; however, an infectious etiology is still being sought. KD has seasonal peaks in the winter and spring months, and focal epidemics occurred in the 1970s and 1980s (115). Treatment with aspirin and intravenous immunoglobulin (IVIG) has reduced the development and severity of coronary artery aneurysms. Other Causes Streptococcus viridans bacteremia can cause generalized erythema. Ehrlichiosis can produce a toxic shock-like syndrome with diffuse erythema. Enteroviral infections, graft versus host disease, and erythroderma may all present with diffuse erythema (8).

VESICULAR, BULLOUS, OR PUSTULAR RASHES Vesicles and bullae refer to small and large blisters. Pustules refer to a vesicle filled with cloudy fluid. The causes of vesiculobullous rashes associated with fever include primary varicella infection, herpes zoster, HSV, smallpox, S. aureus bacteremia, gonococcemia, Vibrio vulnificus, Rickettsia akari, enteroviral infections, parvovirus B19, and HIV infection. Other causes that will not be discussed include folliculitis due to staphylococci, Pseudomonas aeruginosa, and Candida, but these manifestations would not result in admission to a critical care unit. Varicella Zoster Primary infection with varicella (chickenpox) is usually more severe in adults and immunocompromised patients. Although it can be seen year-round, the highest incidence of infection occurs in the winter and spring. The disease presents with a prodrome of fever and malaise one to two days prior to the outbreak of the rash. The rash begins as erythematous macules that quickly develop into vesicles. The characteristic rash is described as ‘‘a dewdrop on a rose petal.’’ The vesicles evolve into pustules that umbilicate and crust. A characteristic of primary varicella is that lesions in all stages may be present at one time (8). Herpes zoster (i.e., shingles) is caused by the reactivation of the varicella-zoster virus (VZV), which lies dormant in the basal root ganglia (116). The incidence of zoster is greatest in older age groups because of a decline in VZV-specific cellmediated immunity. Herpes zoster also occurs more often in immunosuppressed patients, such as transplant recipients (117–119) and HIV-infected patients (120–122). Patients often have a prodrome of fever, malaise, headaches, and dysesthesias that precedes the vesicular eruption by several days (123). The characteristic rash usually affects a single dermatome and begins as an erythematous maculopapular eruption that quickly evolves into a vesicular rash (Fig. 8). The lesions then dry and crust over in 7 to 10 days, with resolution in 14 to 21 days (116). Disseminated herpes zoster is seen in patients with solid-organ transplants, hematological malignancies, and HIV infection (120,121,124–128). Thirty-five percent of patients who

382

Engel et al.

Figure 8 Lower abdomen of a patient with a herpes-zoster outbreak due to varicella-zoster virus. Source: Courtesy of the CDC, Public Health Image Library.

have received bone-marrow transplants have reactivation of VZV, and 50% of these patients develop disseminated herpes zoster (126,129,130). Both immunocompetent and immunocompromised patients can have complications from herpes zoster; however, the risk is greater for immunocompromised patients (131). Complications of herpes zoster include herpes zoster ophthalmalicus (124,132), acute retinal necrosis (133,134), Ramsay Hunt syndrome (135), aseptic meningitis (136), peripheral motor neuropathy (136), myelitis (136,137), encephalitis (136), pneumonitis (131), hepatitis (129), and pancreatitis (126). The diagnosis of primary varicella infection and herpes zoster is often made clinically. Diagnosis of the neurological complications can be made with CSF PCR assays (138,139). Patients with ocular involvement should be seen promptly by an ophthalmologist. Smallpox Smallpox is caused by the variola virus. The last known case of naturally acquired smallpox occurred in Somalia in 1977. The World Health Organization declared that smallpox had been eradicated from the world in 1980 as a result of global vaccination (140,141). The only known repositories for this virus are in Russia and the United States. With the threat of bioterrorism, there is still a remote possibility that this entity would be part of the differential diagnosis of a vesicular rash. Smallpox usually spreads by respiratory droplets, but infected clothing or bedding can also spread disease (142). The incidence of smallpox is highest during the winter and early spring. The pox virus can survive longer at lower temperatures and low levels of humidity (143,144). After a 12-day incubation period, smallpox infection presents with a prodromal phase of acute onset of fever (often greater than 40 C), headaches, and backaches

Fever and Rash in the Critical Care Unit

383

(142). A macular rash develops and progresses to vesicles and then pustules over one to two weeks (145). The rash appears on the face, oral mucosa, and arms first but then gradually involves the whole body. The pustules are 4 to 6 mm in diameter and remain for five to eight days, after which time, they umbilicate and crust. The lesions of smallpox are generally all in the same stage of development (Fig. 9). ‘‘Pock’’ marks are seen in 65% to 80% of survivors. Historically, the case mortality rate was 20% to 50% (142). In the United States, almost nobody under the age of 30 has been vaccinated; therefore this group is largely susceptible to infection. The diagnosis of smallpox is clinically based on the presence of a characteristic rash that is centrifugal in distribution. After fluid from a vesicle is obtained by someone who has received a smallpox vaccine, laboratory confirmation can quickly be made by electron microscopic examination. Definitive identification in the laboratory is accomplished with viral cell culture, PCR, and restriction fragment–length polymorphism analysis (146). Herpes Simplex Virus HSV type 1 (herpes labialis) commonly causes vesicular lesions of the oral mucosa (147). The illness is characterized by the sudden appearance of multiple, often painful, vesicular lesions on an erythematous base. The lesions last for 10 to 14 days. Recurrent infections in the immunocompetent host are usually shorter than the primary infection. In the immunocompromised host, infections can be much more serious. Aside from vesicular eruptions on mucous membranes, the infection can cause keratitis, acute retinal necrosis, hepatitis, esophagitis, pneumonitis, and neurological syndromes (147–156). HSV-1 can cause sporadic cases of encephalitis characterized by rapid onset of fever, headache, seizures, focal neurological signs, and impaired mental function. HSV-1 encephalitis has a high rate of mortality in the absence of treatment (157).

Figure 9 Male Patient with smallpox. Source: Courtesy of the CDC/Barbara Rice, Public Health Image Library.

384

Engel et al.

Diagnosis can infrequently be made by culture; PCR analysis of the CSF has become the gold-standard technique for making the diagnosis (158). S. aureus Bacteremia S. aureus can cause metastatic skin infections that often manifest as pustules (3). The pustular skin eruption due to this organism is often widespread. Risk factors for bacteremia include older age, diabetes, recent surgery, HIV, hemodialysis, neoplasm, neutropenia, and intravenous drug use (159–164). Bacteremia can lead to metastatic complications, such as endocarditis and arthritis. Risk factors for these metastatic complications include underlying valvular heart disease and prosthetic implants. V. vulnificus V. vulnificus is a slightly curved, gram-negative bacillus that is endemic in warm coastal waters around the world. V. vulnificus is the leading cause of seafood-related fatalities in the United States (165). There are reports that virtually all oysters and 10% of crabs harvested in the warmer summer months from the Gulf of Mexico are culture positive for V. vulnificus (166). Consequently, the illness presents mostly between March and November (167). In the United States, most cases occur in states bordering the Gulf of Mexico or those that import oysters from the Gulf States (168). Risk factors for infection include liver disease (most commonly alcoholic), hemachromatosis, HIV infection, steroid use, malignancy, and achlohydria (165). V. vulnificus has been associated with two distinct syndromes: septicemia and wound infection (169,170). A third syndrome of gastrointestinal illness has also been suggested (171). Primary septicemia is a fulminant illness that occurs after the consumption of contaminated raw shellfish. Consumption of raw oysters within 14 days preceding the illness has been reported in 96% of the cases (172). Wound infection occurs after a pre-existing or newly acquired wound is exposed to contaminated seawater. The onset of symptoms is abrupt. The most common presenting signs and symptoms are fever, chills, shock, and secondary bullae (170). Skin lesions are seen in 65% of patients and are an early sign of septicemia. The most characteristic skin manifestation is erythema, followed by a rapid development of indurated plaques. These plaques then become violaceous in color, vesiculate, and then form bullae. The necrotic skin eventually sloughs off, leaving large ulcers (Fig. 10) (173). Gangrene of a limb can develop due to blood-vessel occlusion (174). Diagnosis is aided by clinical presentation and history. The bacteria can be readily cultured from blood and cutaneous lesions (175). A real-time PCR assay has also been reported (176). The mortality rate for septicemia is about 53% and is higher in patients who present with hypotension and leucopenia (177). Median duration from hospitalization to death is about 1.6 days (170). Failure to initiate antibiotics promptly is associated with higher mortality (168). Debridement of involved tissue is usually pursued. R. akari Rickettsialpox, which was first described in 1946 in New York City, is caused by R. akari (178). R. akari infects house mice (Mus muscaulus) and is transmitted to humans by the house mouse–associated mite, Liponyssoides sanguineus (179).

Fever and Rash in the Critical Care Unit

385

Figure 10 Skin Lesions associated with Vibrio vulnificus septicemia in a 75-year-old patient with liver cirrhosis. Source: Courtesy of CDC; From Ref. 173.

Most cases have occurred in large metropolitan areas of the northeastern United States (179,180). Rickettsialpox has an incubation period of 9 to 14 days (181). The initial lesion develops into an eschar at the site of inoculation. Local lymph nodes around the eschar may become enlarged and tender. Approximately one week following the development of the eschar, patients will develop high fever, headaches, malaise, and myalgias. Some patients will have shaking chills and drenching sweats. Thrombocytopenia may also be noted (180). Within three to seven days of the fever, skin eruptions of red macules, papules, and papulovesicles will develop over the body. These lesions number between 20 and 40 and will resolve within a week. The presence of an eschar, the lack of successive crops of vesicles over time, and the presence of thrombocytopenia will help differentiate this entity from VZV infection (180). Diagnosis has been made by comparing acute and convalescent serum antibody titers. Indirect and direct fluorescent antibody tests using anti–Rickesttsia rikettsii antibodies have also been reported (179). The length of the disease course can be reduced with tetracycline, but even untreated patients typically recover without complication (179).

NODULAR RASH A nodule is a palpable, solid, round, or ellipsoidal lesion that may contain inflammatory cells, organisms (fungi and mycobacterium), or cancer cells (5). Nodules usually result from disease in the dermis.

386

Engel et al.

Erythema Nodosum Erythema nodosum is an acute inflammatory process involving the fatty-tissue layer and skin. This condition is more common in woman. There are several causes (Table 8), including infections with streptococci, Chlamydia species, and hepatitis C (182–186). The presentation includes fever, malaise, and arthralgias. The characteristic nodules are painful and tender. The nodules commonly develop over the lower legs, knees, and arms (182). Spontaneous resolution usually occurs within six weeks. Diagnosis is often clinical, but biopsy may be needed in atypical cases. Systemic Fungal Infections The sudden onset of dermal nodules may indicate disseminated candidiasis. Risk factors for disseminated candidiasis include malignancy, neutropenia, antimicrobial therapy, severe burn injuries, intravenous catheters, and systemic steroid administration (187–189). The lesions are raised erythematous papules or nodules that are discrete, firm, and nontender (189–191). Other fungi, such as blastomycosis, histoplasmosis, coccidiodomycosis, and sporotrichosis, can also produce skin nodules (5,192). Patients with AIDS may present with umbilicated nodules that resemble Molluscum contagiousum but are caused by Cryptococcus neoformans.

Table 8 Causes of Erythema Nodosum Infectious Bacterial infections Streptococcus pyogenes Mycobacterium tuberculosis Mycobacterium leprae Cat scratch disease Chlamydia Enteric pathogens (Yersinea, Campylobacter, Salmonella) Rickettsiae Spirochetes (syphilis) Systemic fungal infections Coccidiodes immitis Histoplasma capsulatum Blastomycosis Parasites Amebiasis Giardiasis Ascaris Viral infections Hepatitis B CMV EBV

Non-Infectious Drug reactions Oral contraceptives Antibiotics Hepatitis B vaccine Sulfonamides Systemic disease Systemic lupus erythematosis Ulcerative colitis Crohn’s disease Leukemia Lymphoma Sarcoidosis Idiopathic (55%)

Abbreviations: CMV, cytomegalovirus; EBV, Epstein–Barr virus. Source: Adapted from Ref. 182.

Fever and Rash in the Critical Care Unit

387

Rheumatic Fever Rheumatic fever is a late inflammatory complication of acute group A streptococcal pharyngitis (193,194). Rheumatic fever occurs two to four weeks following the pharyngitis. This disease occurs most frequently in children between the ages of four to nine years. The disease is self-limited, but resulting damage to the heart valves may be chronic and progressive, leading to cardiac decompensation and death. Rheumatic fever is an acute, systemic, febrile illness that can produce a migratory arthritis, carditis, central nervous system deficits, and rash. The diagnosis is based on major and minor criteria (i.e., modified Jones Criteria) (195). The five major criteria are carditis, polyarthritis, chorea, erythema marginatum, and subcutaneous nodules. The three minor criteria are fever, arthralgia, and previous rheumatic fever or rheumatic heart disease. Arthritis is the most frequent and least specific manifestation (196). Large joints are affected most commonly. The arthritis is migratory, with the joints of the lower extremities affected first, followed by those of the upper extremities. Carditis associated with rheumatic fever manifests as pericarditis, myocarditis, and endocarditis, most commonly involving the mitral valve, followed by the aortic valve (197,198). Rheumatic heart disease is a late sequela of acute rheumatic fever, occurring 10 to 20 years after the acute attack, and is the most common cause of acquired valvular disease in the world (199). The mitral valve is most commonly affected with resultant mitral stenosis that often requires surgical correction. Syndenham chorea (chorea minor, St. Vitus’ Dance) is a neurological disorder that manifests as abrupt, purposeless, involuntary movements, muscle weakness, and emotional disturbances (200). The abnormal movements disproportionately affect one side of the body and cease during sleep. Subcutaneous nodules are firm and painless and are seen most often with patients who have carditis (201). The overlying skin is not inflamed. The nodules can be as large as 2 cm and are most commonly located over bony surfaces or near tendons. The nodules may be present for one to four weeks. Erythema marginatum (202) is a pink or faint-red, nonpruritic rash that affects the trunk and proximal limbs and spares the face. Erythema marginatum occurs early in the disease and may persist or recur. The rash is usually only seen in patients with concomitant carditis. The diagnosis of rheumatic fever is supported by evidence of preceding Group A streptococcal infection. Evidence of increased antistreptolysin O antibodies, positive throat culture for Group A beta-hemolytic streptococci, positive rapid-direct Group A streptococcus carbohydrate antigen test, or recent scarlet fever along with the presence of one major and two minor or two major criteria is considered adequate to make the diagnosis.

REFERENCES 1. Cunha BA. Rash and fever in the critical care unit. Crit Care Clin 1998; 14(1):35–53. 2. Kingston ME, Mackey D. Skin clues in the diagnosis of life-threatening infections. Rev Infect Dis 1986; 8(1):1–11. 3. Schlossberg D. Fever and rash. Infect Dis Clin North Am 1996; 10(1):101–110. 4. Sanders CV. Approach to the diagnosis of the patient with fever and rash. In: Sanders CV, Nesbitt LTJ, eds. The skin and Infection: A Color Atlas and Text. Baltimore: Williams and Wilkins, 1995:296–304.

388

Engel et al.

5. Weber DJ, Cohen MS, Rutala WA. The acutely ill patient with fever and rash. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Disease. Philadelphia: Elsevier Churchill Livingstone, 2005:729–746. 6. Garner JS. Hospital Infection Control Practices Advisory Committee. Guideline for isolation precautions in hospitals. Infect Control Hosp Epidemiol 1996; 17:53–80. 7. Centers for Disease Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care facilities. MMWR Recomm Rep 1994; 43:1–132. 8. McKinnon HD Jr., Howard T. Evaluating the febrile patient with a rash. Am Fam Physician 2000; 62(4):804–816. 9. Nesbitt LTJ. Evaluating the patient with a skin infection-General considerations. In: Sanders CV, Nesbitt LTJ, eds. The Skin and Infection: A Color Atlas and Text. Baltimore: Williams and Wilkins, 1995:1–6. 10. Schuchat A, Robinson K, Wenger JD, et al. Bacterial meningitis in the United States in 1995. Active Surveillance Team. N Engl J Med 1997; 337(14):970–976. 11. Yung AP, McDonald MI. Early clinical clues to meningococcemia. MJA 2003; 178: 134–137. 12. van Deuren M, Brandtzaeg P, van der Meer JW. Update on meningococcal disease with emphasis on pathogenesis and clinical management. Clin Microbiol Rev 2000; 13(1): 144–166. 13. Pollard AJ, Britto J, Nadel S, et al. Emergency management of meningococcal disease. Arch Dis Child 1999; 80(3):290–296. 14. Mclean S, Caffey J. Endemic purpuric meningococcus bacteremia in early life: the diagnostic value of smears from purpuric lesions. Am J Dis Child 1931; 42:1053–1074. 15. Hill WR, Kinney TD. The cutaneous lesions in acute meningococcemia: a clinical and pathological study. JAMA 1947; 134:513–518. 16. Faust SN, Levin M, Harrison OB, et al. Dysfunction of endothelial protein C activation in severe meningococcal sepsis. N Engl J Med 2001; 345(6):408–416. 17. Bernhard WG, Jordan AC. Purpuric lesions in meningococcal infections: diagnosis from smears and cultures of the purpuric lesions. J Lab Clin Med 1944; 29:273–281. 18. van Deuren M, van Dijke BJ, Koopman RJ, et al. Rapid diagnosis of acute meningococcal infections by needle aspiration or biopsy of skin lesions. BMJ 1993; 306(6887): 1229–1232. 19. Gregory B, Tron V, Ho VC. Cyclic fever and rash in a 66-year-old woman. Chronic meningococcemia. Arch Dermatol 1992; 128(12):1645–1648. 20. Ploysangam T, Sheth AP. Chronic meningococcemia in childhood: case report and review of the literature. Pediatr Dermatol 1996; 13(6):483–487. 21. Masters EJ, Olson GS, Weiner SJ, et al. Rocky Mountain spotted fever: a clinician’s dilemma. Arch Intern Med 2003; 163(7):769–774. 22. Sexton DJ, Kaye KS. Rocky mountain spotted fever. Med Clin North Am 2002; 86(2): 351–360. 23. Thorner AR, Walker DH, Petri WA Jr. Rocky mountain spotted fever. Clin Infect Dis 1998; 27(6):1353–1359. 24. Jacobs RF. Human monocytic ehrlichiosis: similar to Rocky Mountain spotted fever but different. Pediatr Ann 2002; 31(3):180–184. 25. Holman RC, Paddock CD, Curns AT, et al. Analysis of risk factors for fatal Rocky Mountain Spotted Fever: evidence for superiority of tetracyclines for therapy. J Infect Dis 2001; 184(11):1437–1444. 26. Kirkland KB, Marcom PK, Sexton DJ, et al. Rocky Mountain spotted fever complicated by gangrene: report of six cases and review. Clin Infect Dis 1993; 16(5):629–634. 27. Rangel-Frausto MS. The epidemiology of bacterial sepsis. Infect Dis Clin North Am 1999; 13(2):299–312. 28. Annane D, Bellissant E, Cavaillon JM. Septic shock. Lancet 2005; 365(9453):63–78.

Fever and Rash in the Critical Care Unit

389

29. Alberti C, Brun-Buisson C, Goodman SV, et al. Influence of systemic inflammatory response syndrome and sepsis on outcome of critically ill infected patients. Am J Respir Crit Care Med 2003; 168(1):77–84. 30. Mylonakis E, Calderwood SB. Infective endocarditis in adults. N Engl J Med 2001; 345(18):1318–1330. 31. Strom BL, Abrutyn E, Berlin JA, et al. Risk factors for infective endocarditis: oral hygiene and nondental exposures. Circulation 2000; 102(23):2842–2848. 32. Manoff SB, Vlahov D, Herskowitz A, et al. Human immunodeficiency virus infection and infective endocarditis among injecting drug users. Epidemiology 1996; 7(6):566–570. 33. Brown PD, Levine DP. Infective endocarditis in the injection drug user. Infect Dis Clin North Am 2002; 16(3):645–665. 34. Crawford MH, Durack DT. Clinical presentation of infective endocarditis. Cardiol Clin 2003; 21(2):159–166. 35. Barr J, Danielsson D. Septic gonococcal dermatitis. Br Med J 1971; 1(747):482–485. 36. Holmes KK, Counts GW, Beaty HN. Disseminated gonococcal infection. Ann Intern Med 1971; 74(6):979–993. 37. Kerle KK, Mascola JR, Miller TA. Disseminated gonococcal infection. Am Fam Physician 1992; 45(1):209–214. 38. Handsfield HH. Disseminated gonococcal infection. Clin Obstet Gynecol 1975; 18(1): 131–142. 39. Abu-Nassar H, Hill N, Fred HL, et al. Cutaneous manifestations of gonococcemia. A review of 14 cases. Arch Intern Med 1963; 112:731–737. 40. Krol-van Straaten MJ, Landheer JE, de Maat CE. Capnocytophaga canimorsus (formerly DF-2) infections: review of the literature. Neth J Med 1990; 36(5–6):304–309. 41. Lion C, Escande F, Burdin JC. Capnocytophaga canimorsus infections in human: review of the literature and case report. Eur J Epidemiol 1996; 12(5):521–533. 42. Kullberg BJ, Westendorp RG, ’t Wout JW, et al. Purpura fulminans and symmetrical peripheral gangrene caused by Capnocytophaga canimorsus (formerly DF-2) septicemia— a complication of dog bite. Medicine 1991; 70(5):287–292. 43. Guha-Sapir D, Schimmer B. Dengue fever: new paradigms for a changing epidemiology. Emerg Themes Epidemiol 2005; 2(1):1–10. 44. Calisher CH. Persistent emergence of dengue. Emerg Infect Dis 2005; 11(5):738–739. 45. Shirtcliffe P, Cameron E, Nicholson KG, et al. Don’t forget dengue! Clinical features of dengue fever in returning travelers. J Roy Coll Physicians Lond 1998; 32(3):235–237. 46. Kalayanarooj S, Vaughn DW, Nimmannitya S, et al. Early clinical and laboratory indicators of acute dengue illness. J Infect Dis 1997; 176(2):313–321. 47. Kao CL, King CC, Chao DY, et al. Laboratory diagnosis of dengue virus infection: current and future perspectives in clinical diagnosis and public health. J Microbiol Immunol Infect 2005; 38(1):5–16. 48. Walker DH. Tick-transmitted infectious diseases in the United States. Annu Rev Public Health 1998; 19:237–269. 49. Taege AJ. Tick trouble: overview of tick-borne diseases. Cleve Clin J Med 2000; 67(4): 241, 245–249. 50. Lyme disease—United States, 1999. Morb Mortal Wkly Rep 2001; 50(10):181–185. 51. Gayle A, Ringdahl E. Tick-borne diseases. Am Fam Physician 2001; 64(3):461–466. 52. Steere AC, Coburn J, Glickstein L. The emergence of Lyme disease. J Clin Invest 2004; 113(8):1093–1101. 53. Edlow JA. Lyme disease and related tick-borne illnesses. Ann Emerg Med 1999; 33(6): 680–693. 54. Roujeau JC, Stern RS. Severe adverse cutaneous reactions to drugs. N Engl J Med 1994; 331(19):1272–1285. 55. Roujeau JC. Clinical heterogeneity of drug hypersensitivity. Toxicology 2005; 209(2): 123–129.

390

Engel et al.

56. Pruksachatkunakorn C, Schachner LA. Erythema Multiforme. www.emedicine.com/ derm/topic137.htm, 2005:1–9. 57. Assier H, Bastuji-Garin S, Revuz J, et al. Erythema multiforme with mucous membrane involvement and Stevens-Johnson syndrome are clinically different disorders with distinct causes. Arch Dermatol 1995; 131(5):539–543. 58. Bastuji-Garin S, Rzany B, Stern RS, et al. Clinical classification of cases of toxic epidermal necrolysis, Stevens-Johnson syndrome, and erythema multiforme. Arch Dermatol 1993; 129(1):92–96. 59. Hayes EB, Komar N, Nasci RS, et al. Epidemiology and transmission dynamics of West Nile Virus disease. Emerg Infect Dis 2005; 11(8):1167–1173. 60. Centers for Disease Control and Prevention (CDC). West Nile virus activity—United States, 2005. Morb Mortal Wkly Rep 2005; 54(31):769–770. 61. O’Leary DR, Marfin AA, Montgomery SP, et al. The epidemic of West Nile virus in the United States, 2002. Vector Borne Zoonotic Dis 2004; 4(1):61–70. 62. Petersen LR, Hayes EB. Westward ho?—The spread of West Nile virus. N Engl J Med 2004; 351(22):2257–2259. 63. Zeller HG, Schuffenecker I. West Nile virus: an overview of its spread in Europe and the Mediterranean basin in contrast to its spread in the Americas. Eur J Clin Micro Infect Dis 2004; 23(3):147–156. 64. Hayes EB, O’Leary DR. West Nile virus infection: a pediatric perspective. Pediatrics 2004; 113(5):1375–1381. 65. Kumar D, Prasad GV, Zaltzman J, et al. Community-acquired West Nile virus infection in solid-organ transplant recipients. Transplantation 2004; 77(3):399–402. 66. Stramer SL, Fang CT, Foster GA, et al. West Nile virus among blood donors in the United States, 2003 and 2004. N Engl J Med 2005; 353(5):451–459. 67. Diamond MS, Shrestha B, Mehlhop E, et al. Innate and adaptive immune responses determine protection against disseminated infection by West Nile encephalitis virus. Viral Immunol 2003; 16(3):259–278. 68. Mostashari F, Bunning ML, Kitsutani PT, et al. Epidemic West Nile encephalitis, New York, 1999: results of a household-based seroepidemiological survey. Lancet 2001; 358(9278):261–264. 69. Watson JT, Pertel PE, Jones RC, et al. Clinical characteristics and functional outcomes of West Nile fever. Ann Intern Med 2004; 141(5):360–365. 70. Hayes EB, Sejvar JJ, Zaki, et al. Virology, pathology, and clinical manifestations of West Nile Virus. Emerg Infect Dis 2005; 11(8):1174–1179. 71. Pepperell C, Rau N, Krajden S, et al. West Nile virus infection in 2002: morbidity and mortality among patients admitted to hospital in south-central Ontario. CMAJ 2003; 168(11):1399–1405. 72. Sejvar JJ, Haddad MB, Tierney BC, et al. Neurologic manifestations and outcome of West Nile virus infection. JAMA 2003; 290(4):511–515. 73. Sejvar JJ, Bode AV, Marfin AA, et al. West Nile virus-associated flaccid paralysis. Emerg Infect Dis 2005; 11(7):1021–1027. 74. Tilley PA, Zachary GA, Walle R, et al. West Nile virus detection and commercial assays. Emerg Infect Dis 2005; 11(7):1154–1155. 75. Martin DA, Biggerstaff BJ, Allen B, et al. Use of immunoglobulin M cross-reactions in differential diagnosis of human flaviviral encephalitis infections in the United States. Clin Diagn Lab Immunol 2002; 9(3):544–549. 76. Reich HL, Crawford GH, Pelle MT, et al. Group B streptococcal toxic shock-like syndrome. Arch Dermatol 2004; 140(2):163–166. 77. Todd J, Fishaut M, Kapral F, et al. Toxic-shock syndrome associated with phage-group-I Staphylococci. Lancet 1978; 2(8100):1116–1118. 78. Stevens DL, Tanner MH, Winship J, et al. Severe group A streptococcal infections associated with a toxic shock-like syndrome and scarlet fever toxin A. N Engl J Med 1989; 321(1):1–7.

Fever and Rash in the Critical Care Unit

391

79. Hajjeh RA, Reingold A, Weil A, et al. Toxic shock syndrome in the United States: surveillance update, 1979–1996. Emerg Infect Dis 1999; 5(6):807–810. 80. Reduced incidence of menstrual toxic-shock syndrome—United States, 1980–1990. MMWR Morb Mortal Wkly Rep 1990; 39(25):421–423. 81. Bergdoll MS, Crass BA, Reiser RF, et al. A new staphylococcal enterotoxin, enterotoxin F, associated with toxic-shock-syndrome Staphylococcus aureus isolates. Lancet 1981; 1(8228):1017–1021. 82. Schlievert PM. Staphylococcal enterotoxin B and toxic-shock syndrome toxin-1 are significantly associated with non-menstrual TSS. Lancet 1986; 1(8490):1149–1150. 83. Schlievert PM, Shands KN, Dan BB, et al. Identification and characterization of an exotoxin from Staphylococcus aureus associated with toxic-shock syndrome. J Infect Dis 1981; 143(4):509–516. 84. Lee VT, Chang AH, Chow AW. Detection of staphylococcal enterotoxin B among toxic shock syndrome (TSS)- and non-TSS-associated Staphylococcus aureus isolates. J Infect Dis 1992; 166(4):911–915. 85. Lehn N, Schaller E, Wagner H, et al. Frequency of toxic shock syndrome toxin- and enterotoxin-producing clinical isolates of Staphylococcus aureus. Eur J Clin Micro Infect Dis 1995; 14(1):43–46. 86. Barrett JA, Graham DR. Toxic shock syndrome presenting as encephalopathy. J Infect 1986; 12(3):276–278. 87. Chesney RW, Chesney PJ, Davis JP, et al. Renal manifestations of the staphylococcal toxic-shock syndrome. Am J Med 1981; 71(4):583–588. 88. Herzer CM. Toxic shock syndrome: broadening the differential diagnosis. J Am Board Fam Pract 2001; 14(2):131–136. 89. McKenna UG, Meadows JA III, Brewer NS, et al. Toxic shock syndrome, a newly recognized disease entity. Report of 11 cases. Mayo Clin Proc 1980; 55(11):663–672. 90. Chesney PJ, Davis JP, Purdy WK, et al. Clinical manifestations of toxic shock syndrome. JAMA 1981; 246(7):741–748. 91. Andrews MM, Parent EM, Barry M, et al. Recurrent nonmenstrual toxic shock syndrome: clinical manifestations, diagnosis, and treatment. Clin Infect Dis 2001; 32(10): 1470–1479. 92. Reingold AL, Dan BB, Shands KN, et al. Toxic-shock syndrome not associated with menstruation. A review of 54 cases. Lancet 1982; 1(8262):1–4. 93. Reingold AL, Hargrett NT, Dan BB, et al. Nonmenstrual toxic shock syndrome: a review of 130 cases. Ann Intern Med 1982; 96(6 Pt 2):871–874. 94. Defining the group A streptococcal toxic shock syndrome. Rationale and consensus definition. The Working Group on Severe Streptococcal Infections. JAMA 1993; 269(3): 390–391. 95. Manders SM. Toxin-mediated streptococcal and staphylococcal disease. J Am Acad Dermatol 1998; 39(3):383–398. 96. Ladhani S. Recent developments in staphylococcal scalded skin syndrome. Clin Micro Infect 2001; 7(6):301–307. 97. Adesiyun AA, Lenz W, Schaal KP. Exfoliative toxin production by Staphylococcus aureus strains isolated from animals and human beings in Nigeria. Microbiologica 1991; 14(4): 357–362. 98. Dancer SJ, Noble WC. Nasal, axillary, and perineal carriage of Staphylococcus aureus among women: identification of strains producing epidermolytic toxin. J Clin Path 1991; 44(8):681–684. 99. Ladhani S, Poston SM, Joannou CL, et al. Staphylococcal scalded skin syndrome: exfoliative toxin A (ETA) induces serine protease activity when combined with A431 cells. Acta Paediatr 1999; 88(7):776–779. 100. Papageorgiou AC, Plano LR, Collins CM, et al. Structural similarities and differences in Staphylococcus aureus exfoliative toxins A and B as revealed by their crystal structures. Protein Sci 2000; 9(3):610–618.

392

Engel et al.

101. Prevost G, Couppie P, Monteil H. Staphylococcal epidermolysins. Curr Opin Infect Dis 2003; 16(2):71–76. 102. Lyell A. Toxic epidermal necrolysis (the scalded skin syndrome): a reappraisal. Br J Dermatol 1979; 100(1):69–86. 103. Melish ME. Staphylococci, streptococci and the skin: review of impetigo and staphylococcal scalded skin syndrome. Semin Dermatol 1982; 1:101–109. 104. Beers B, Wilson B. Adult staphylococcal scalded skin syndrome. Int J Dermatol 1990; 29(6):428–429. 105. Moss C, Gupta E. The Nikolsky sign in staphylococcal scalded skin syndrome. Arch Dis Child 1998; 79(3):290. 106. Gemmell CG. Staphylococcal scalded skin syndrome. J Med Micro 1995; 43(5):318–327. 107. Cribier B, Piemont Y, Grosshans E. Staphylococcal scalded skin syndrome in adults. A clinical review illustrated with a new case. J Am Acad Dermatol 1994; 30(2 Pt 2):319–324. 108. Goldberg NS, Ahmed T, Robinson B, et al. Staphylococcal scalded skin syndrome mimicking acute graft-vs.-host disease in a bone marrow transplant recipient. Arch Dermatol 1989; 125(1):85–87. 109. Ladhani S, Robbie S, Garratt RC, et al. Development and evaluation of detection systems for staphylococcal exfoliative toxin A responsible for scalded-skin syndrome. J Clin Microbiol 2001; 39(6):2050–2054. 110. Gentilhomme E, Faure M, Piemont Y, et al. Action of staphylococcal exfoliative toxins on epidermal cell cultures and organotypic skin. J Dermatol 1990; 17(9):526–532. 111. Leyden JJ, Gately LE III. Staphylococcal and streptococcal infections. In: Sanders CV, Nesbitt LTJ, eds. The Skin and Infection: A Color Atlas and Text. Baltimore: Williams and Wilkins, 1995:27–38. 112. Kawasaki T. Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desquamation of the fingers and toes in children. Arerugi 1967; 16(3):178–222. 113. Burns JC, Glode MP. Kawasaki syndrome. Lancet 2004; 364(9433):533–544. 114. Royle J, Burgner D, Curtis N. The diagnosis and management of Kawasaki disease. J Pediatr Child Health 2005; 41(3):87–93. 115. Burns JC, Cayan DR, Tong G, et al. Seasonality and temporal clustering of Kawasaki syndrome. Epidemiology 2005; 16(2):220–225. 116. Morgan R, King D. Shingles: a review of diagnosis and management. Hospital Med (Lond) 1998; 59(10):770–776. 117. Feldman S, Hughes WT, Kim HY. Herpes zoster in children with cancer. Am J Dis Child 1973; 126(2):178–184. 118. Feldhoff CM, Balfour HH Jr., Simmons RL, et al. Varicella in children with renal transplants. J Pediatr 1981; 98(1):25–31. 119. Patti ME, Selvaggi KJ, Kroboth FJ. Varicella hepatitis in the immunocompromised adult: a case report and review of the literature. Am J Med 1990; 88(1):77–80. 120. Veenstra J, van Praag RM, Krol A, et al. Complications of varicella zoster virus reactivation in HIV-infected homosexual men. AIDS 1996; 10(4):393–399. 121. Buchbinder SP, Katz MH, Hessol NA, et al. Herpes zoster and human immunodeficiency virus infection. J Infect Dis 1992; 166(5):1153–1156. 122. Gershon AA, Mervish N, LaRussa P, et al. Varicella-zoster virus infection in children with underlying human immunodeficiency virus infection. J Infect Dis 1997; 176(6): 1496–1500. 123. Straus SE, Ostrove JM, Inchauspe G, et al. NIH conference. Varicella-zoster virus infections. Biology, natural history, treatment, and prevention. Ann Intern Med 1988; 108(2): 221–237. 124. Galil K, Choo PW, Donahue JG, et al. The sequelae of herpes zoster. Arch Intern Med 1997; 157(11):1209–1213. 125. Levitsky J, Kalil A, Meza JL, et al. Herpes zoster infection after liver transplantation: a case-control study. Liver Transpl 2005; 11(3):320–325.

Fever and Rash in the Critical Care Unit

393

126. Rogers SY, Irving W, Harris A, et al. Visceral varicella zoster infection after bone marrow transplantation without skin involvement and the use of PCR for diagnosis. Bone Marrow Transplant 1995; 15(5):805–807. 127. Veenstra J, Krol A, van Praag RM, et al. Herpes zoster, immunological deterioration and disease progression in HIV-1 infection. AIDS 1995; 9(10):1153–1158. 128. Rusthoven JJ, Ahlgren P, Elhakim T, et al. Varicella-zoster infection in adult cancer patients. A population study. Arch Intern Med 1988; 148(7):1561–1566. 129. Tojimbara T, So SK, Cox KL, et al. Fulminant hepatic failure following varicella-zoster infection in a child. A case report of successful treatment with liver transplantation and perioperative acyclovir. Transplantation 1995; 60(9):1052–1053. 130. Verdonck LF, Cornelissen JJ, Dekker AW, et al. Acute abdominal pain as a presenting symptom of varicella-zoster virus infection in recipients of bone marrow transplants. Clin Infect Dis 1993; 16(1):190–191. 131. Fleisher G, Henry W, McSorley M, et al. Life-threatening complications of varicella. Am J Dis Child 1981; 135(10):896–899. 132. Pavan-Langston D. Herpes zoster ophthalmicus. Neurology 1995; 45(12 suppl 8): S50–S51. 133. Hellinger WC, Bolling JP, Smith TF, et al. Varicella-zoster virus retinitis in a patient with AIDS-related complex: case report and brief review of the acute retinal necrosis syndrome. Clin Infect Dis 1993; 16(2):208–212. 134. Culbertson WW, Blumenkranz MS, Pepose JS, et al. Varicella zoster virus is a cause of the acute retinal necrosis syndrome. Ophthalmology 1986; 93(5):559–569. 135. Adour KK. Otological complications of herpes zoster. Ann Neurol 1994; 35(suppl): S62–S64. 136. Elliott KJ. Other neurological complications of herpes zoster and their management. Ann Neurol 1994; 35(suppl):S57–S61. 137. Au WY, Hon C, Cheng VC, et al. Concomitant zoster myelitis and cerebral leukemia relapse after stem cell transplantation. Ann Hematol 2005; 84(1):59–60. 138. Burke DG, Kalayjian RC, Vann VR, et al. Polymerase chain reaction detection and clinical significance of varicella-zoster virus in cerebrospinal fluid from human immunodeficiency virus-infected patients. J Infect Dis 1997; 176(4):1080–1084. 139. Cinque P, Bossolasco S, Vago L, et al. Varicella-zoster virus (VZV) DNA in cerebrospinal fluid of patients infected with human immunodeficiency virus: VZV disease of the central nervous system or subclinical reactivation of VZV infection? Clin Infect Dis 1997; 25(3):634–639. 140. Breman JG, Arita I. The confirmation and maintenance of smallpox eradication. N Engl J Med 1980; 303(22):1263–1273. 141. Henderson DA. Smallpox eradication. Public Health Rep 1980; 95(5):422–426. 142. Dixon CW. Smallpox. London: J & A Churchill Ltd., 1962. 143. Huq F. Effect of temperature and relative humidity on variola virus in crusts. Bull World Health Org 1976; 54(6):710–712. 144. Harper GJ. The influence of environment on the survival of airborne virus particles in the laboratory. Arch Gesamte Virusforsch 1963; 13:64–71. 145. Breman JG, Henderson DA. Diagnosis and management of smallpox. N Engl J Med 2002; 346(17):1300–1308. 146. Henderson DA, Inglesby TV, Bartlett JG, et al. Smallpox as a biological weapon: medical and public health management. Working Group on Civilian Biodefense. JAMA 1999; 281(22):2127–2137. 147. Corey L, Spear PG. Infections with herpes simplex viruses (1). N Engl J Med 1986; 314(11):686–691. 148. Frederick DM, Bland D, Gollin Y. Fatal disseminated herpes simplex virus infection in a previously healthy pregnant woman. A case report. J Reprod Med 2002; 47(7): 591–596.

394

Engel et al.

149. Hillard P, Seeds J, Cefalo R. Disseminated herpes simplex in pregnancy: two cases and a review. Obstet Gynecol Surv 1982; 37(7):449–453. 150. Johnson JR, Egaas S, Gleaves CA, et al. Hepatitis due to herpes simplex virus in marrow-transplant recipients. Clin Infect Dis 1992; 14(1):38–45. 151. Kusne S, Schwartz M, Breinig MK, et al. Herpes simplex virus hepatitis after solid organ transplantation in adults. J Infect Dis 1991; 163(5):1001–1007. 152. Liesegang TJ. Herpes simplex virus epidemiology and ocular importance. Cornea 2001; 20(1):1–13. 153. Muller SA, Herrmann EC Jr., Winkelmann RK. Herpes simplex infections in hematologic malignancies. Am J Med 1972; 52(1):102–114. 154. Priya K, Mahalakshmi B, Malathi J, et al. Prevalence of herpes simplex virus, varicella zoster virus and cytomegalovirus in HIV-positive and HIV-negative patients with viral retinitis in India. Eur J Clin Micro Infect Dis 2004; 23(11):857–858. 155. Remeijer L, Osterhaus A, Verjans G. Human herpes simplex virus keratitis: the pathogenesis revisited. Ocular Immunol Inflamm 2004; 12(4):255–285. 156. Siegal FP, Lopez C, Hammer GS, et al. Severe acquired immunodeficiency in male homosexuals, manifested by chronic perianal ulcerative herpes simplex lesions. N Engl J Med 1981; 305(24):1439–1444. 157. Tyler KL. Herpes simplex virus infections of the central nervous system: encephalitis and meningitis, including Mollaret’s. Herpes 2004; 11(suppl 2):57A–64A. 158. DeBiasi RL, Kleinschmidt-DeMasters BK, Weinberg A, et al. Use of PCR for the diagnosis of herpes virus infections of the central nervous system. J Clin Virol 2002; 25(suppl 1):S5–S11. 159. Gopal AK, Fowler VG Jr., Shah M, et al. Prospective analysis of Staphylococcus aureus bacteremia in non-neutropenic adults with malignancy. J Clin Oncol 2000; 18(5): 1110–1115. 160. Gottlieb GS, Fowler VG Jr., Kong LK, et al. Staphylococcus aureus bacteremia in the surgical patient: a prospective analysis of 73 postoperative patients who developed Staphylococcus aureus bacteremia at a tertiary care facility. J Am Coll Surg 2000; 190(1):50–57. 161. Marr KA, Kong L, Fowler VG, et al. Incidence and outcome of Staphylococcus aureus bacteremia in hemodialysis patients. Kidney Int 1998; 54(5):1684–1689. 162. Petti CA, Fowler VG Jr. Staphylococcus aureus bacteremia and endocarditis. Cardiol Clin 2003; 21(2):219–233. 163. Shah MA, Sanders L, Lanclos K, et al. Staphylococcus aureus bacteremia in patients with neutropenia. South Med J 2002; 95(7):782–784. 164. Tumbarello M, de Gaetano DK, Tacconelli E, et al. Risk factors and predictors of mortality of methicillin-resistant Staphylococcus aureus (MRSA) bacteremia in HIV-infected patients. J Antimicrob Chemo 2002; 50(3):375–382. 165. Mitra AK. Vibrio vulnificus infection: epidemiology, clinical presentations, and prevention. South Med J 2004; 97(2):118–119. 166. Ruple AD, Cook DW. Vibrio vulnificus indicator bacteria in shellstock and commercially processed oysters. J Food Prot 1992; 55:667–671. 167. Hill MK, Sanders CV. Localized and systemic infection due to Vibrio species. Infect Dis Clin North Am 1987; 1(3):687–707. 168. Klontz KC, Lieb S, Schreiber M, et al. Syndromes of Vibrio vulnificus infections. Clinical and epidemiologic features in Florida cases, 1981–1987. Ann Intern Med 1988; 109(4): 318–323. 169. Chiang SR, Chuang YC. Vibrio vulnificus infection: clinical manifestations, pathogenesis, and antimicrobial therapy. J Microb Immun Infect 2003; 36(2):81–88. 170. Haq SM, Dayal HH. Chronic liver disease and consumption of raw oysters: a potentially lethal combination—a review of Vibrio vulnificus septicemia. Am J Gastro 2005; 100(5): 1195–1199.

Fever and Rash in the Critical Care Unit

395

171. Levine WC, Griffin PM, Gulf Coast Vibrio Working Group. Vibrio infections on the Gulf Coast: results of first year of regional surveillance. J Infect Dis 1993; 167(2):479–483. 172. Howard RJ, Bennett NT. Infections caused by halophilic marine Vibrio bacteria. Ann Surg 1993; 217(5):525–530. 173. Hsueh PR, Lin CY, Tang HJ, et al. Vibrio vulnificus in Taiwan. Emerg Infect Dis 2004; 10(8):1363–1368. 174. Wickboldt LG, Sanders CV. Vibrio vulnificus infection. Case report and update since 1970. J Am Acad Dermatol 1983; 9(2):243–251. 175. Koenig KL, Mueller J, Rose T. Vibrio vulnificus. Hazard on the half shell. West J Med 1991; 155(4):400–403. 176. Chen CY, Wu KM, Chang YC, et al. Comparative genome analysis of Vibrio vulnificus, a marine pathogen. Genome Res 2003; 13(12):2577–2587. 177. Morris JG Jr. Vibrio vulnificus—a new monster of the deep? Ann Intern Med 1988; 109(4):261–263. 178. Greenberg M, Pellitteri OJ, Jellison WL. Rickettsialpox-a newly recognized rickettsial disease. Am J Public Health 1947; 37:860–868. 179. Kass EM, Szaniawski WK, Levy H, et al. Rickettsialpox in a New York City hospital, 1980 to 1989. N Engl J Med 1994; 331(24):1612–1617. 180. Krusell A, Comer JA, Sexton DJ. Rickettsialpox in North Carolina: a case report. Emerg Infect Dis 2002; 8(7):727–728. 181. Brettman LR, Lewin S, Holzman RS, et al. Rickettsialpox: report of an outbreak and a contemporary review. Medicine 1981; 60(5):363–372. 182. Cribier B, Caille A, Heid E, et al. Erythema nodosum and associated diseases. A study of 129 cases. Int J Dermatol 1998; 37(9):667–672. 183. Blomgren SE. Conditions associated with erythema nodosum. NY State J Med 1972; 72(18):2302–2304. 184. James DG. Erythema nodosum. Br Med J 1961; 5229:853–857. 185. Puavilai S, Sakuntabhai A, Sriprachaya-Anunt S, et al. Etiology of erythema nodosum. J Med Assoc Thai 1995; 78(2):72–75. 186. Vesey CM, Wilkinson DS. Erythema nodosum: a study of seventy cases. Br J Dermatol 1959; 71(4):139–155. 187. Maksymiuk AW, Thongprasert S, Hopfer R, et al. Systemic candidiasis in cancer patients. Am J Med 1984; 77(4D):20–27. 188. Bodey GP. Fungal infection and fever of unknown origin in neutropenic patients. Am J Med 1986; 80(5C):112–119. 189. Bodey GP. Candidiasis in cancer patients. Am J Med 1984; 77(4D):13–19. 190. Balandran L, Rothschild H, Pugh N, et al. A cutaneous manifestation of systemic candidiasis. Ann Intern Med 1973; 78(3):400–403. 191. Jacobs MI, Magid MS, Jarowski CI. Disseminated candidiasis. Newer approaches to early recognition and treatment. Arch Dermatol 1980; 116(11):1277–1279. 192. Radentz WH. Opportunistic fungal infections in immunocompromised hosts. J Am Acad Dermatol1920(6):989–1003. 193. Hahn RG, Knox LM, Forman TA. Evaluation of poststreptococcal illness. Am Fam Physician 2005; 71(10):1949–1954. 194. Stollerman GH. Rheumatic fever in the 21st century. Clin Infect Dis 2001; 33(6): 806–814. 195. Guidelines for the diagnosis of rheumatic fever. Jones Criteria, 1992 update. Special Writing Group of the Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the Council on Cardiovascular Disease in the Young of the American Heart Association. JAMA 1992; 268(15):2069–2073. 196. Feinstein AR, Spagnuolo M. The clinical patterns of acute rheumatic fever: a reappraisal. Medicine 1962; 41:279–305. 197. Stollerman GH. Rheumatic fever. Lancet 1997; 349(9056):935–942.

396

Engel et al.

198. Hilario MO, Andrade JL, Gasparian AB, et al. The value of echocardiography in the diagnosis and follow up of rheumatic carditis in children and adolescents: a 2 year prospective study. J Rheumatol 2000; 27(4):1082–1086. 199. McCallum AH. Natural history of rheumatic fever and rheumatic heart disease. Ten-year report of a co-operative clinical trial of A.C.T.H., Cortisone, and aspirin. Br Med J 1965; 5462:607–613. 200. Swedo SE. Sydenham’s chorea. A model for childhood autoimmune neuropsychiatric disorders. JAMA 1994; 272(22):1788–1791. 201. Baldwin JS, Kerr JM, Kuttner AG, et al. Observations on rheumatic nodules over a 30-year period. J Pediatr 1960; 56:465–470. 202. Burke JB. Erythema marginatum. Arch Dis Child 1955; 30(152):359–365.

19

Tropical Infections in the Critical Care Unita David R. Tribble Department of Enteric Diseases, Infectious Diseases Directorate, Naval Medical Research Institute, Silver Spring, Maryland, U.S.A.

Kenneth F. Wagner Independent Consultant, Infectious Diseases and Tropical Medicine, Islamorada, Florida, U.S.A.

EPIDEMIOLOGY OF INFECTIONS IN INTERNATIONAL TRAVELERS International travel brings a world of experiences and opportunities but also carries a degree of health risks that can often be prevented or better managed if appropriate pretravel preparation is undertaken. Retrospective case series of traveler health risks have documented the wide range of health problems that may be associated with international travel (1). Both noninfectious and infectious diseases are well represented with the predominant causes of death being accidents (motor vehicle and drowning) and cardiovascular related in younger and older travelers, respectively (2). Infections account for significant morbidity and mortality both during and after international travel, particularly into developing tropical regions (1,3). There is a predominance of enteric transmission as the most common route of infection with the usual clinical syndrome being traveler’s diarrhea (1,3). Food- and water-borne infections such as typhoid fever, cholera, hepatitis A, and uncommonly agents of traveler’s diarrhea occasionally result in critical illness in travelers. The most common life-threatening infection in returning travelers is malaria, and this will be emphasized in this chapter. The determinants of infectious risk and potential etiologies in a returning traveler are based on relative incidence of the infection, regional distribution, traveler predisposition, and specific aspects of the traveler’s itinerary or activities. Steffen et al. calculated incidences (per 100,000 travelers) of various infectious diseases in returning European travelers from developing regions (1). Severe diarrhea was the most common illness (12,998) followed by diarrhea with fever (1940), acute respiratory tract infection with fever (1261), giardiasis (660), hepatitis (446), amebiasis (427), gonorrhea (330), and malaria (97). Notably absent in this study were

a

The opinions and assertions contained herein are those of the authors and do not reflect the official policy of the Department of Navy, Department of Defense, or the U.S. Government. 397

398

Tribble and Wagner

typhoid and cholera, which have been reported in other retrospective surveys at rates of approximately 5/100,000 and 100 parasites/mL), negative results requiring microscopic confirmation, and cost concerns (32). Rapid diagnosis of malaria is critical due to the semiquantitative relationship between level of parasitemia and mortality: 500,000 parasites/mL ¼ 72% mortality (33). A qualified laboratory is a necessity based on the lack of clinical predictors. Standard methods are thick and thin peripheral blood Giemsa or Wright stained smears obtained serially (every six to eight hours over 24 hours usually a minimum of three smears). Appropriate expertise and timely referral to a reference laboratory if necessary is critical. The management of malaria is reliant on prompt recognition and initiation of effective therapy with a blood schizonticide to rapidly reduce parasitemia (34). After recognizing that the patient has malaria, the next steps are to differentiate species between P. falciparum and nonfalciparum malaria. If the traveler with falciparum malaria reports travel limited to chloroquine-susceptible regions (Central America, Haiti, and possibly limited areas in the Middle East), then chloroquine therapy can be used (intravenous and oral available). Given the widespread chloroquine resistance, monotherapy should only be used in areas where treatment efficacy has been recently demonstrated and not for severe malaria (25,35). Severe malaria should be managed with parenteral antimalarial therapy given the potential for erratic absorption through the gastrointestinal tract (36). Recommended treatment modalities currently available are from the cinchona alkaloid class [quinine dihydrochloride (intravenous or intramuscular) or quinidine gluconate (intravenous)] or artemisinin derivatives [artesunate (intravenous) or artemether (intramuscular)] (34,36). The artemisinins more rapidly clear parasitemia with equivalent efficacy to intravenous quinine even with rectal delivery (37–40). In the U.S., the only licensed drug for intravenous therapy is quinidine gluconate often combined with a second blood schizonticide for radical cure (such as tetracycline) (Table 2) (41). Loading doses are used to rapidly attain effective drug levels with the exception of quinine or

Tropical Infections in the Critical Care Unit

401

quinidine loading in patients who have been receiving quinine, mefloquine, or quinidine (34). Monitoring of the QRS and QTc intervals is needed during quinidine therapy due to the potential for systemic hypotension or QT prolongation (42). Exchange transfusion is occasionally used for severe malaria when parasitemia levels exceed 10% or if the patient has altered mental status, nonvolume overload pulmonary edema, or renal complications (43). It is usually continued until the parasite load is 5%), (ii) altered level of consciousness (cerebral malaria; r/o hypoglycemia), (iii) circulatory shock (typically due to gram-negative sepsis), Table 2 Severe Malaria Treatment Regimens Available in the United Statesa Drug Parenteral administration All severe malaria Quinidine gluconate plus one of the below

Doxycyclinebor

Clindamycin

a

Adult dosage

Pediatric dosage

10 mg salt/kg load (max 600 mg) in normal saline over 1–2 hr then 0.02 mg salt/kg/min continuous infusion for at least 48–72 hr or 24 mg salt/kg load over 4 hr, followed by 12 mg salt/kg infused over 4 hr every 8 hr starting 8 hr after loading dose for at least 48–72 hr. Switch to PO quinine sulfate 650 mg q8h when parasite density 20%), Glascow Coma Score 8, blood glucose 65 years or Updated Oct 6, 2005. 3. Narasimhan M, Posner AJ, DePaol VA, et al. Intensive care in patients with HIV infection in the era of highly active antiretroviral therapy. Chest 2004; 125:1800–1804. 4. Morris A, Masur H, Huang L. Current issues in critical care of the human immunodeficiency virus-infected patient. Crit Care Med 2006; 34:42–49. 5. Palacios R, Santos J, Camino X, et al. Treatment-limiting toxicities associated with nucleoside analogue reverse transcriptase inhibitor therapy: a prospective, observational study. Curr Ther Res Clin Exp 2005; 66:117–129. 6. Gertner E, Thurn JR, Williams DN, et al. Zidovudine-associated myopathy. Am J Med 1989; 86:814–818. 7. Kakuda T. Pharmacology of nucleoside and nucleoside reverse transcriptase inhibitorinduced mitochondrial toxicity. Clin Ther 2000; 22:685–708. 8. Gerschenson M, Brinkman K. Mitochondrial dysfunction in AIDS and its treatment. Mitochondrion 2004; 4:763–777. 9. Stryer K, ed. Biochemistry. New York: WH Freeman and Co., 1988:397–426.

HIV/AIDS in the Critical Care Unit

429

10. Dagan T, Sable C, Bray J, Gerschenson M. Mitochondrial dysfunction and antiretroviral nucleoside analog toxicities: what is the evidence? Mitochondrion 2002; 1:397–412. 11. Arnoult D, Violett L, Petit F, et al. HIV-1 triggers mitochondrial death. Mitochondrion 2004; 4:255–260. 12. Hoschele D. Cell culture models for the investigation of NRTI-induced mitochondrial toxicity. Relevance for the prediction of clinical toxicity. Toxicology In Vitro. 2006; 20: 535–54. 13. Lai KK, Gang DL, Zawacki JK, Cooley TP. Fulminant hepatic failure associated with 20 , 30 dideoxyinosine. Ann Intern Med 1991; 115:283–284. 14. Ogedegbe A-E, Thomas DL, Diehl AM. Hyperlactataemia syndromes associated with HIV therapy. Lancet Infect Dis 2003; 3:329–337. 15. Calza L, Manfredi R, Chiodo F. Hyperlactateaemia and lactic acidosis in HIV-infected patients receiving antiretroviral therapy. Clin Nutrition 2005; 24:5–15. 16. Moyle GJ, Datta D, Mandalia S, et al. Hyperlactataemia and lactic acidosis during antiretroviral therapy: relevance, reproducibility and possible risk factors. AIDS 2002; 16: 1341–1349. 17. John M, Mallal S. Hyperlactataemia syndrome in people with HIV infection. Curr Opin Infect Dis 2002; 15:23–29. 18. Falco V, Rodriquez D, Ribera E, et al. Severe nucleoside-associated lactic acidosis in human immunodeficiency virus-infected patients: report of 12 cases and review of the literature. Clin Infect Dis 2002; 34:838–846. 19. Verma A, Roland M, Jayaweere D, Kett D. Fulminat neuropathy and lactic acidosis associated with nucleoside analog therapy. Neurology 1999; 53:1365–1369. 20. Simpson D, Estanislao L, Evans S, et al. HIV-associated neuromuscular weakness syndrome. AIDS 2004; 18:1403–1412. 21. Estanislao L, Thomas D, Simpson D. HIV neuromuscular disease and mitochondrial function. Mitochondrion 2004; 4:131–139. 22. Bjornsson E, Olsson R. Suspected drug-induced liver fatalities reported to the WHO database. Dig Liver Dis 2006; 38:33–38. 23. Cattelan AM, Erne E, Stalino A, et al. Severe hepatic failure related to nevirapine treatment. Clin Infect Dis 1999; 29:455–456. 24. Prakash M, Poreddy V, Tiyyagura L, Bonacini M. Jaundice and hepatocellular damage associated with nevirapine therapy. Am J Gastroenterol 2001; 96:1571–1574. 25. Pollard RB, Robinson P, Dransfield K. Safety profile of nevirapine, a nonnucleoside reverse transcriptase inhibitor for the treatment of human immunodeficiency virus infection. Clin Therap 1998; 20:1071–1092. 26. Gonzales de Requena D, Nunez M, Jimenez-Nacher I, Soriano V. Liver toxicity caused by nevirapine. AIDS 2005; 19:621–623. 27. Reisler K. High hepatotoxicity rate seen among HAART patients. AIDS Alert 2001; 16:118–119. 28. De Maat MMR, Mathot RAA, Veldkam AI, et al. Hepatotoxicity following nevirapinecontaining regimens in HIV-1-infected individuals. Pharmacological Res 2002; 46:295–300. 29. Joao EC, Calvet GA, Menezes JA, et al. Nevirapine toxicity in a cohort of HIV-1-infected pregnant women. Am J Obstet Gynecol 2006; 194:199–202. 30. Centers for Disease Control and Prevention: Serious adverse events attributed to nevirapine regimens for postexposure prophylaxis after HIV exposures – Worldwide, 1997–2000. MMWR 2001; 49:1153–1156. 31. Food and Drug Administration Center for Drug Evaluation and Research: FDA public health advisory for nevirapine (Viramune). http://www.fda.gov/cder/drug/advisory/ Nevirapine.htm. 32. Hetherington S, McGuirk S, Powell G, et al. Hypersensitivity reactions during therapy with the nucleoside reverse transcriptase inhibitor abacavir. Clin Ther 2001; 23:1603–1614. 33. Clay PG. The abacavir hypersensitivity reaction: a review. Clin Therap 2001; 24:1502–1514.

430

Lutwick

34. Loeliger AE, Steel H, McGuirk S, et al. The abacavir hypersensitivity reaction and interruptions in therapy. AIDS 2001; 15:1325–1326. 35. Mallal S, Nolan D, Witt C, et al. Association between presence of HLA-B5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet 2002; 359:727–732. 36. Hetherington S, Hughes AR, Mosteller M, et al. Genetic variations in HLA-B region and hypersensitivity reactions in abacavir. Lancet 2002; 359:1121–1122. 37. Martin AM, Nolan D, Gaudieri S, et al. Predisposition to abacavir hypersensitivity conferred by HLA-B5701 and a haplotypic Hsp70-Hom variant. Proc Natl Acad Sci USA 2004; 101:4180–4185. 38. King D, Tomkins S, Waters A, et al. Intracellular cytokines may model immunoregulation of abacavir hypersensitivity in HIV-infected subjects. J Allergy Clin Immunol 2005; 115: 1081–1087. 39. Carr A, Tindall B, Penny R, Cooper DA. Patterns of multiple-drug hypersensitivities in HIV-infected patients. AIDS 1993; 7:1532–1533. 40. Nunn P, Kibuga D, Gathua S, et al. Cutaneous hypersensitivity reactions due to thiacetazone in HIV-1 seropositive patients treated for tuberculosis. Lancet 1991; 337:627–630. 41. Fagot J-P, Mockenhaupt M, Bouwes-Bavinck J-N, et al. Nevirapine and the risk of StevenJohnson syndrome or toxic epidermal necrolysis. AIDS 2001; 15:1843–1848. 42. Barner A, Myers M. Nevirapine and rashes. Lancet 1998; 351:1133. 43. Anton P, Soriano V, Jimenez-Nacher I, et al. Incidence of rash and discontinuation of nevirapine using two different escalating initial doses. AIDS 1999; 13:524–525. 44. Claes P, Wintzen M, Allard S, et al. Nevirapine-induced toxic epidermal necrolysis and toxin hepatitis treated successfully with a combination of intravenous immunoglobulins and N-acetylcysteine. Eur J Intern Med 2004; 15:255–258. 45. Barbaro G, Fisher SD, Luipshultz SE. Pathogenesis of HIV-associated cardiovascular complications. Lancet Infect Dis 2001; 1:115–124. 46. Lipshultz SE. Dilated cardiomyopathy in HIV-infected patients. N Engl J Med 1998; 339:1153–1155. 47. Silva-Cardoso J, Moura B, Martins L, et al. Pericardial involvement in human immunodeficiency virus infection. Chest 1999; 115:418–422. 48. Walli R, Herfort O, Michl GM, et al. Treatment with protease inhibitors associated with peripheral insulin resistance and impaired glucose tolerance in HIV-1-infected patients. AIDS 1998; 12:F167–F173. 49. Behrens G, Dejam A, Schmidt H, et al. Impaired glucose tolerance, beta cell function, and lipid metabolism in HIV patients under treatment with protease inhibitors. AIDS 1999; 13:F63–F70. 50. Tsiodras S, Mantzonos C, Hammer S, Samore M. Effects of protease inhibitors on hyperglycemics, hyperlipidemia, and lipodystrophy: a 5-year cohort study. Arch Intern Med 2000; 160:2050–2056. 51. Hatano H, Miller KD, Yoder CP, et al. Metabolic and anthropometric consequences of interruption of highly active antiretroviral therapy. AIDS 2000; 14:1935–1942. 52. Fisher SD, Miller TL, Lipshultz SE. Impact of HIV and highly active antiretroviral therapy on leukocyte adhesions molecules, arterial inflammation, dyslipidemia, and atherosclerosis. Atherosclerosis 2006; 185:1–11. 53. Holmberg SD, Moorman AC, Williamson JM, et al. Protease inhibitors and cardiovascular outcomes in patients with HIV-1. Lancet 2002; 360:1747–1748. 54. Mary-Kraus M, Cotte L, Simon A, et al. The Clinical Epidemiology Group from the French Hospital Database. Increased risk of myocardial infarction with duration of protease inhibitor therapy in HIV-infected men. AIDS 2003; 17:2479–2486. 55. Coplan PM, Nikas A, Japour A, et al. Incidence of myocardial infarction in randomized clinical trials of protease inhibitor-based antiretroviral therapy: an analysis of four different protease inhibitors. AIDS Res Hum Retroviruses 2003; 19:449–455.

HIV/AIDS in the Critical Care Unit

431

56. Klein D, Hurley LB, Quesenberry CP, Sidney S. Do protease inhibitors increase the risk for coronary heart disease in patients with HIV-1 infection? J Acquir Immune Defic Syndr 2002; 30:471–477. 57. Bozzette SA, Ake CF, Tam HK, et al. Cardiovascular and cerebrovascular events in patients treated for human immunodeficiency virus infection. N Engl J Med 2003; 348:702–710. 58. de Saint Marftin, Vandhuick O, Guillo P, et al. Premature atherosclerosis in HIV-positive patient and cumulated time of exposure to antiretroviral therapy (SHIVA study). Atherosclerosis. 2006; 185:361–367.

21 Infections in Cirrhosis in the Critical Care Unit Laurel C. Preheim Departments of Medicine, Medical Microbiology and Immunology, Creighton University School of Medicine, University of Nebraska College of Medicine, and VA Medical Center, Omaha, Nebraska, U.S.A.

INTRODUCTION Cirrhosis is characterized by fibrosis of the hepatic parenchyma with regenerative nodules surrounded by scar tissue. It can result from a variety of chronic, progressive liver diseases. The clinical manifestations vary widely from asymptomatic disease (up to 40% of patients) to fulminant liver failure. Cirrhosis is a major cause of morbidity worldwide. In the United States, cirrhosis has an estimated prevalence of 360 per 100,000 population and accounts for approximately 30,000 deaths annually. The majority of cases in the United States are due to alcoholic liver disease or chronic infection with hepatitis B or C viruses. Infection is a common complication of cirrhosis (reviewed in Refs. 1–4). A Danish death registry study (5) examined long-term survival and cause-specific mortality in 10,154 patients with cirrhosis between 1982 and 1993. The results revealed an increased risk of dying from respiratory infection (fivefold), from tuberculosis (15-fold), and other infectious diseases (22-fold) when compared to the general population. In a recent prospective study (6), 20% of cirrhotic patients admitted to the hospital developed an infection while hospitalized. The mortality among patients with infection was 20% compared to 4% mortality in those who remained uninfected. Of patients admitted to the critical care unit, 41% became infected. The most common bacterial infections seen in cirrhotic patients are urinary tract infections (12–29%), spontaneous bacterial peritonitis (SBP) (7–23%), respiratory tract infections (6–10%), and primary bacteremia (4–11%) (7). The increased susceptibility to bacterial infections among cirrhotic patients is related to impaired hepatocyte and phagocytic cell function as well as the consequences of parenchymal destruction (portal hypertension, ascites, and gastroesophageal varices). It should be noted that the usual signs and symptoms of infection may be subtle or absent in individuals who have advanced liver disease. Thus a high index of suspicion is required to ensure that infections are not overlooked in this patient population, especially in those who are hospitalized. Occasionally fever may be 433

434

Preheim

due to cirrhosis itself (8), but this must be a diagnosis of exclusion made only when appropriate diagnostic tests, including cultures, have been unrevealing. ROLE OF THE LIVER IN HOST DEFENSE AGAINST INFECTION The liver plays an important role in host defense against infection. Cirrhosis can adversely affect a number of these host defenses. The mechanisms identified in human and experimental animal studies include depression of reticuloendothelial system clearance of organisms from the bloodstream (9); impairment of chemotaxis, phagocytosis, and intracellular killing by polymorphonuclear leukocytes (PMNL) and monocytes (10–12); reduction in serum bactericidal activity and opsonic activity (13,14); depression of serum complement (15–17); dysregulation of cytokine synthesis and metabolism (18); and reduced protective efficacy of type-specific antibody (19) and granulocyte colony-stimulating factor (20). CLASSIFICATION OF LIVER DISEASE SEVERITY Patients who have cirrhosis are at increased risk for both community-acquired and nosocomial infections, the majority of which are bacterial. The incidence of infection is highest for patients with the most severe liver disease (6,21–23). Accurate assessment for risk of infection is dependent upon proper classification of the extent of liver disease. The Child–Pugh scoring system of liver disease severity (24) is based upon five parameters: serum bilirubin, serum albumin, prothrombin time, ascites, and encephalopathy. A total score is derived from the sum of the points for each of these five parameters. Patients with chronic liver disease are placed in one of three classes (A, B, or C). Despite having some limitations, the modified Child–Pugh scoring system continues to be used by many clinicians to assess the risk of mortality in patients with cirrhosis (Table 1). SPONTANEOUS BACTERIAL PERITONITIS Pathogenesis SBP is the infection of ascitic fluid with no identifiable abdominal source for the infection. SBP is perhaps the most characteristic bacterial infection in cirrhosis, Table 1 Modified Child–Pugh Classification of Liver Disease Severity Points assigned Parameter

1

Ascites Encephalopathy Bilirubin (mg/dL) Albumin (mg/L) Prothrombin time (seconds increased)

None None 3.5 1–3 Total score 5–6 7–9 10–15

2 Slight Grade 1–2 2.0–3.0 2.8–3.5 4–6

3 Moderate/severe Grade 3–4 >3.0 6.0 Child–Pugh class A B C

Infections in Cirrhosis in the Critical Care Unit

435

occurring in as many as 20% to 30% of cirrhotic patients who are admitted to the hospital with ascites (6,21,23). SBP occurs when normally sterile ascitic fluid is colonized following an episode of transient bacteremia. Aerobic gram-negative bacilli, especially Escherichia coli, cause approximately 75% of SBP infections. Aerobic gram-positive cocci, including Streptococcus pneumoniae, Enterococcus faecalis, other streptococci, and Staphylococcus aureus, are responsible for most other SBP cases (25,26). Because enteric bacteria predominate in SBP it is thought that the gut is the major source of organisms for this infection. Several mechanisms have been proposed to explain the movement of organisms from the intestinal lumen to the systemic circulation (reviewed in Ref. 1). Cirrhosis-induced depression of the hepatic reticuloendothelial system (RES) impairs the liver’s filtering function, allowing bacteria to pass from the bowel lumen to the bloodstream via the portal vein. Cirrhosis also is associated with a relative increase in aerobic gram-negative bacilli in the jejunum. A decrease in mucosal blood flow due to acute hypovolemia or drug-induced splanchnic vasoconstriction may compromise the intestinal barrier to enteric flora, thereby increasing the risk of bacteremia. Finally, bacterial translocation may occur with movement of enteric organisms from the gut lumen through the mucosa to the intestinal lymphatics. From there bacteria can travel through the lymphatic system and enter the bloodstream via the thoracic duct. It is assumed that SBP caused by nonenteric organisms also is due to bacteremia secondary to another site of infection with subsequent seeding of the peritoneum and ascitic fluid (Fig. 1). Decreased opsonic activity of ascitic fluid also increases the risk of SBP in patients with cirrhosis. Immunoglobulin, complement, and fibronectin are important opsonins in ascitic fluid, and patients with low protein concentrations in their ascitic fluid are especially predisposed to SBP (27,28). Patients with ascitic fluid protein concentrations below 1 g/dL have a sevenfold increase in the incidence of SBP when compared to patients with higher protein concentrations in ascites (27). Other risk factors have been associated with SBP, including gastrointestinal bleeding, fulminant hepatic failure, and invasive procedures such as the placement of peritoneovenous shunts for the treatment of ascites. An elevated bilirubin level also is correlated with a high risk of peritonitis in patient with cirrhosis (28).

Enteric bacteria (primarily coliforms)

Non-enteric bacteria

Portal vein

Bacterial translocation to lymphatics

Impaired reticuloendothelial system function Portosystemic shunting

Bacteremia Impaired reticuloendothelial system function

Seeding of peritoneal fluid Decreased ascitic fluid opsonic activity

Spontaneous bacterial peritonitis

Figure 1 Pathogenic mechanisms underlying spontaneous bacterial peritonitis. Source: Adapted from Ref. 1.

436

Preheim

Diagnosis Classic signs and symptoms of peritonitis, including fever, chills, abdominal pain, and increasing ascites may or may not be present in cirrhotic patients who have SBP. Abdominal symptoms may be absent in up to one-third of cases. Patients with SBP may present with encephalopathy, gastrointestinal bleeding, or increasing renal insufficiency. Therefore a high index of suspicion must be maintained in all cases of cirrhotic patients who have ascites and are acutely ill. A diagnostic paracentesis must be performed on all patients suspected to have SBP. A cell count of greater than 250 PMNL/mm3 of ascitic fluid is highly suggestive of infection. Gram stain of centrifuged ascitic fluid will reveal organisms in approximately 30% of cases. The fluid should be cultured both aerobically and anaerobically. Inoculating some fluid directly into blood culture bottles increases the yield of positive cultures. But this nonquantitative culture technique also increases the risk of false positives if any skin flora contaminant is introduced into the blood culture bottle at the bedside. As indicated previously, aerobic gram-negative enteric bacilli are the most frequent isolates from ascitic fluid cultures in SBP. Anaerobes are uncommon causes of SBP, and their presence in ascitic fluid should raise suspicions for bowel perforation. If ascitic fluid cultures yield polymicrobial flora, Candida albicans (or other yeast), or Bacteroides fragilis one should suspect a secondary peritonitis caused by an acute abdominal infection. Treatment Historically SBP has been a severe, frequently fatal infection. In the past few decades mortality rates have dropped from over 90% in the 1970s to the current 20%– 40% mortality for patients who have their first diagnosis of SBP. Earlier detection and treatment and the use of non-nephrotoxic antibiotics has contributed to the increased short-term survival. The most common causes of death in patients with SBP are liver failure, gastrointestinal bleeding, and renal failure. One of the greatest threats to long-term survival is recurrence of SBP, which can occur in 70% of patients (29). Previously aminoglycosides, alone or in combination with beta-lactam antibiotics, were widely used to treat SBP. However, the risk of aminoglycoside nephrotoxicity in cirrhotic patients has limited the usefulness of this class of agents (30). Expandedspectrum cephalosporins are active against most of the strains of enteric gram-negative pathogens that cause SBP. Cefotaxime has been shown effective in a number of trials with regimens of 2 g administered every eight hours for five days (26) or 2 g every 12 hours for a mean of nine days (31). In a more recent study (32) 24 of 33 cirrhotic patients (73%) with SBP had clinical and bacteriologic cures after receiving 1 g of ceftriaxone every 12 hours for five days. With prolonged treatment using ceftriaxone or with a change to another antibiotic according to susceptibility, SBP resolved in seven of the nine patients who had not responded by day 5 of therapy. Study patients had an overall hospital mortality of only 12%. The authors concluded that antibiotic therapy for SBP can be discontinued if the polymorphonuclear differential count in ascitic fluid is less than 250 cells/mm3 on day 5 of treatment (32). Other parenteral antibiotics that have been reported effective for the treatment of SBP include aztreonam (500 mg every eight hours) (33), cefonicid (2 g every 12 hours) (34), and amoxicillin-clavulanic acid (35). Several small trials have involved the use of oral antibiotics. These included intravenous followed by oral therapy with amoxicillin-clavulanic acid (36) or ciprofloxacin (37) and oral ofloxacin (38). While some experts recommend that patients with moderate symptoms and a positive

Infections in Cirrhosis in the Critical Care Unit

437

response to a short course of intravenous antibiotics could benefit from therapy with oral fluoroquinolones (29), others have found the supporting evidence to be inconclusive (39). Deterioration of renal function is the most sensitive predictor of in-hospital mortality in patients with SBP (40). In a randomized, multicenter comparative study, patients with SBP who received intravenous albumin for plasma volume expansion plus cefotaxime had less renal impairment and significantly lower mortality (22%) than those receiving cefotaxime alone (41%) (41). The dose of albumin used in this study was 1.5 g/kg of body weight at the time of diagnosis followed by 1 g/kg on day 3. Prophylaxis The use of prophylactic antibiotics decreases the incidence and mortality of bacterial infections, including SBP, in patients who are hospitalized with cirrhosis and ascites (7). Cirrhotic patients who recover from SBP also are at increased risk of subsequent episodes. The one-year probability of recurrence of SBP in this population has been estimated to approach 70% (42). Antibiotics reported effective in preventing SBP have included trimethoprim/sulfamethoxazole (43) and, more commonly, fluoroquinolones such as norfloxacin, ofloxacin, and ciprofloxacin (7,44–46). A major concern regarding repeated or prolonged courses of antibiotic prophylaxis is selection for resistant bacterial pathogens. There are a growing number of recent reports of the development of SBP or other infections caused by fluoroquinolone-resistant organisms, including E. coli, Pseudomonas spp., and methicillin-resistant S. aureus (MRSA), in cirrhotic patients on fluoroquinolone prophylaxis (7,47,48). Thus the use of prophylactic antibiotics should be restricted to patients at greatest risk of SBP, weighing the increased risk of inducing resistant bacteria against the benefits of preventing infection. URINARY TRACT INFECTIONS Urinary tract infections account for 25%–40% of infections in hospitalized cirrhotic patients (21,23,49). The majority of these patients have asymptomatic bacteriuria, but approximately one-third have symptomatic infections (23). The incidence of significant bacteriuria (>105 colony forming units/mL) is higher in women than in men and does not correlate with the severity of the underlying liver disease or with the age of the patient (49). The presence of an indwelling urinary catheter increases the risk of infection. The most common pathogens are E. coli and other aerobic gramnegative coliforms. Asymptomatic bacteriuria does not require treatment, particularly in patients with an indwelling urinary catheter. A urine culture should be obtained on any cirrhotic patient suspected to have a urinary tract infection. Antibiotic therapy, when indicated, should be guided by microbiologic susceptibility testing of the urinary isolate. Antibiotic options for empiric therapy of symptomatic infections include fluoroquinolones or expanded-spectrum penicillins or cephalosporins. Indwelling urinary catheters should be removed as soon as possible to reduce the risk of infection.

BACTEREMIA Cirrhosis predisposes patients to systemic bloodstream infections due to intrahepatic blood shunting and impaired bacterial clearance from the portal blood. Bacteremia

438

Preheim

has been reported to occur in approximately 9% of hospitalized cirrhotic patients (50) and accounts for 20% of the infections diagnosed during their hospital stay (23). The incidence of bacteremia increases with the severity of liver disease. The most commonly identified sources of bacteremia have been SBP, urinary tract infections, pneumonia, soft tissue infections, and biliary tract infections (50,51). The pathogens identified in blood cultures from bacteremic patients mirror those responsible for the primary source infections. E. coli, Klebsiella pneumoniae, Aeromonas hydrophila, and other enteric gram-negative aerobes are common causes of bacteremic infections. Most gram-positive bacteremias are due to S. aureus, S. pneumoniae, or other aerobic streptococcal species. Bloodstream infection is associated with a poor prognosis despite appropriate antibiotic therapy. Mortality rates commonly exceed 50% (50,52). Poor outcome is independent of the type of bacteremia (52), but in-hospital mortality has been correlated with the absence of fever, an elevated serum creatinine, and marked leukocytosis (51). Cirrhotic patients with suspected bacteremia should receive empiric therapy directed against the most common gram-negative and gram-positive pathogens in this setting. Antibiotic selection should take into consideration local microbial susceptibility patterns. Usual therapeutic options would include expanded-spectrum cephalosporins, piperacillin/tazobactam, or a fluoroquinolone such as levofloxacin or moxifloxacin. Cirrhotic patients who undergo endoscopic procedures for gastrointestinal hemorrhage or transhepatic procedures are at increased risk of bacteremia. Endoscopic variceal sclerotherapy or band ligation for bleeding esophageal varices is associated with a reported risk of bacteremia ranging from 5% to 30% (53–55). Although the bacteremia associated with these procedures may be brief, cirrhotic patients are susceptible to infections from transient bacteremia. Gastrointestinal hemorrhage itself is an independent risk factor for bacteremia and other infections in cirrhotic patients. Antibiotic administration has been shown to reduce infectious complications and mortality in cirrhotic patients who are hospitalized for gastrointestinal hemorrhage (56–59). Antibiotic prophylaxis is recommended for all cirrhotic inpatients with gastrointestinal bleeding (60,61). Fluoroquinolone antibiotics were used in most trials with a median treatment duration of seven days.

PNEUMONIA Respiratory tract infections account for approximately 20% of the infectious diseases that are diagnosed in hospitalized cirrhotic patients (21,23,62). S. pneumoniae continues to rank first among bacterial pathogens causing community-acquired pneumonia in adults (63). Chronic liver disease has long been recognized as a risk factor for bacteremic pneumococcal pneumonia (64). The mortality rate for pneumococcal bacteremia in cirrhotic patients may exceed 50% despite appropriate antibiotic therapy (65). Other organisms commonly responsible for community-acquired pneumonia include Mycoplasma pneumoniae, Chlamydia pneumoniae, Legionella pneumophila, and Haemophilus influenzae. Cirrhosis has been associated with an increased risk of severe Acinetobacter baumannii community-acquired pneumonia (66). Sputum and blood samples should be obtained for appropriate diagnostic studies, including Gram stain (sputum) and cultures (sputum and blood). Appropriate empiric therapy while awaiting the results of cultures and other tests would include an expanded-spectrum cephalosporin plus a macrolide or a beta-lactam/betalactamase-inhibitor plus a macrolide or a fluoroquinolone (67).

Infections in Cirrhosis in the Critical Care Unit

439

Table 2 Risk Factors for Nosocomial Pneumonia Due to Resistant Bacteria Antimicrobial therapy in preceding 90 days Current hospital stay 5 days High frequency of antibiotic resistance in the community or hospital unit Hospitalization 2 days in preceding 90 days Residence in nursing home or extended care facility Home infusion therapy (including antibiotics) Chronic dialysis within 30 days Home wound care Family member with multidrug-resistant pathogen Immunosuppressive disease and/or therapy Source: Adapted from Ref. 67.

Hospital-acquired pneumonia may be caused by a wide variety of bacteria. Common pathogens include aerobic gram-negative bacilli, such as Pseudomonas aeruginosa, E. coli, K. pneumoniae, Serratia marcescens, Enterobacter species, Proteus species, and Acinetobacter species. S. aureus and S. pneumoniae predominate among gram-positive pathogens, and the incidence of MRSA nosocomial pneumonia is increasing. A number of risk factors have been identified for nosocomial pneumonia caused by multidrug-resistant bacteria (Table 2) (68). Recommended initial empiric antibiotic therapy for nosocomial pneumonia in patients with no risk factors for multidrug-resistant pathogens or P. aeruginosa would be ceftriaxone or a fluoroquinolone or ampicillin/sulbactam or ertapenem. Patients with any risk factors listed in Table 2 or with onset of nosocomial pneumonia after four days of hospitalization are more likely to have infection due to multidrug-resistant pathogens. Initial empiric therapy in such cases should include an antipseudomonal cephalosporin (e.g., cefepime) or antipseudomonal carbapenem (e.g., imipenem) or piperacillin/tazobactam plus an antipseudomonal fluoroquinolone (ciprofloxacin or levofloxacin) plus vancomycin or linezolid if MRSA risk factors are present or there is a high incidence locally (68). Due to increased risks of aminoglycoside-induced nephrotoxicity and ototoxicity, the use of these agents should be avoided in cirrhotic patients if possible (30). OTHER INFECTIONS Vibrio Infections Vibrio bacteria are gram-negative halophilic inhabitants of marine and estuarine environments. Typical infections caused by these organisms include gastroenteritis, wound infections, and septicemia. Infection usually occurs following consumption of contaminated food or water or by cutaneous inoculation through wounds. The most common pathogens include Vibrio cholerae, V. parahaemolyticus, and V. vulnificus. Preexisting liver disease is a major risk factor for Vibrio infections and has been associated with a fatal outcome in both wound infections and primary septicemia (69). V. vulnificus, the most virulent of the noncholera vibrios, can rapidly invade the bloodstream from the gastrointestinal tract. Classic clinical features of V. vulnificus sepsis include the abrupt onset of chills and fever followed by hypotension with subsequent development of disseminated skin lesions within 36 hours of onset. The skin lesions

440

Preheim

progress to hemorrhagic vesicles or bullae and then to necrotic ulcers (70). This syndrome is highly associated with a history of consuming raw oysters. The mortality rate exceeds 50%. Recommended antibiotic therapy includes using an expanded-spectrum cephalosporin plus a tetracycline (e.g., cefotaxime or ceftazidime plus doxycycline) or a fluoroquinolone (e.g., ciprofloxacin) (70).

Endocarditis Infective endocarditis is a relatively unusual complication of cirrhosis. In the past E. coli and S. pneumoniae were commonly implicated in these infections. More recent studies have identified S. aureus as the most common pathogen along with other gram-positive bacteria such as the viridans streptococci and Enterococcus species (71,72). Streptococcus bovis biotypes [recently reclassified as Streptococcus gallolyticus (S. bovis I), Streptococcus lutetiensis (S. bovis II/1), and Streptococcus pasteurianus (S. bovis II/2)] are emerging as another important cause of bacteremia and endocarditis in patients with chronic liver disease (73,74). Endocarditis caused by S. bovis is commonly associated with bivalvular involvement and a high rate of embolic events.

Spontaneous Bacterial Empyema Spontaneous bacterial empyema is an infection of a preexisting hydrothorax in cirrhotic patients. Although the majority of these patients have ascites, the presence of ascites is not a prerequisite for spontaneous bacterial empyema. SBP is present in approximately half of patients who develop empyema. The most common causes of spontaneous bacterial empyema include E. coli, K. pneumoniae, streptococci, including Enterococcus species, and S. bovis. A diagnostic thoracentesis is recommended in patients with cirrhosis who develop pleural effusions and signs and symptoms of infection (75).

REFERENCES 1. Navasa M, Rimola A, Rode´s J. Bacterial infections in liver disease. Semin Liver Dis 1997; 17:323–333. 2. Navasa M, Rode´s J. Bacterial infections in cirrhosis. Liver Int 2004; 24:277–280. 3. Johnson DH, Cunha BA. Infections in cirrhosis. Infect Dis Clin N Am 2001; 15:363–371. 4. Vilstrup H. Cirrhosis and bacterial infections. Romanian J Gastroenterol 2003; 12:297–302. 5. Sørensen HT, Thulstrup AM, Mellemkjar L, et al. Long-term survival and cause-specific mortality in patients with cirrhosis of the liver: a nationwide cohort study in Denmark. J Clin Epidemiol 2003; 56:88–93. 6. Descheˆnes M, Villeneuve J. Risk factors for the development of bacterial infections in hospitalized patients with cirrhosis. Am J Gastroenterol 1999; 94:2193–2197. 7. Soares-Weiser K, Brezis M, Tur-Kaspa R, et al. Antibiotic prophylaxis of bacterial infections in cirrhotic inpatients: a meta-analysis of randomized controlled clinical trials. Scand J Gastroenterol 2003; 38:193–200. 8. Singh N, Yu VL, Wagener MM, et al. Cirrhotic fever in the 1990s: a prospective study with clinical implications. Clin Infect Dis 1997; 24:1135–1138. 9. Rimola A, Soto R, Bory F, et al. Reticuloendothelial system phagocytic activity in cirrhosis and its relation to bacterial infections and prognosis. Hepatology 1984; 4:53–58. 10. Rajkovic IA, Williams R. Abnormalities of neutrophil phagocytosis, intracellular killing, and metabolic activity in alcoholic cirrhosis and hepatitis. Hepatology 1986; 6:252–262.

Infections in Cirrhosis in the Critical Care Unit

441

11. Gentry MJ, Snitily MU, Preheim LC. Phagocytosis of Streptococcus pneumoniae measured in vitro and in vivo in a rat model of carbon tetrachloride-induced liver cirrhosis. J Infect Dis 1995; 171:350–355. 12. Gentry MJ, Snitily MU, Preheim LC. Decreased uptake and killing of Streptococcus pneumoniae within the lungs of cirrhotic rats. Immunol Infect Dis 1996; 6:43–47. 13. Fierer J, Finley F. Serum bactericidal activity against Escherichia coli in patients with cirrhosis of the liver. J Clin Invest 1979; 63:912–921. 14. Lister PD, Mellencamp MA, Preheim LC. Serum-sensitive Escherichia coli multiply in cirrhotic serum. J Lab Clin Med 1992; 120:633–638. 15. Mellencamp MA, Preheim LC. Pneumococcal pneumonia in a rat model of cirrhosis: effects of cirrhosis on pulmonary defense mechanisms against Streptococcus pneumoniae. J Infect Dis 1991; 163:102–108. 16. Homann C, Varming K, Hogasen K, et al. Acquired C3 deficiency in patients with alcoholic cirrhosis predisposes to infection and increased mortality. Gut 1997; 40:544–549. 17. Alcantara RB, Preheim LC, Gentry MJ. The role of pneumolysin’s complement-activating activity during pneumococcal bacteremia in cirrhotic rats. Infect Immun 1999; 67:2862–2866. 18. Baudouin B, Roucloux I, Crusiaux A, et al. Tumor necrosis factor a and interleukin 6 plasma levels in infected cirrhotic patients. Gastroenterology 1993; 104:1492–1497. 19. Preheim LC, Mellencamp MA, Snitily MU, Gentry MJ. Effect of cirrhosis on the production and efficacy of pneumococcal capsular antibody in a rat model. Am Rev Respir Dis 1992; 146:1054–1058. 20. Preheim LC, Snitily MU, Gentry MJ. Effects of granulocyte colony-stimulating factor in cirrhotic rats with pneumococcal pneumonia. J Infect Dis 1996; 174:225–228. 21. Caly WR, Strauss E. A prospective study of bacterial infections in patients with cirrhosis. J Hepatol 1993; 18:353–358. 22. Yoshida H, Hamada T, Inuzuka S, et al. Bacterial infections in cirrhosis, with and without hepatocellular carcinoma. Am J Gastroenterol 1993; 88:2067–2071. 23. Borzio M, Salerno F, Piantoni L, et al. Bacterial infection in patients with advanced cirrhosis: a multicentre prospective study. Digest Liver Dis 2001; 33:41–48. 24. Pugh RN, Murray-Lyon IM, Dawson JL, et al. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973; 60:646–649. 25. Rimola A, Navasa M, Arroyo V. Experience with cefotaxime in the treatment of spontaneous bacterial peritonitis in cirrhosis. Diagn Microbiol Infect Dis 1995; 22:141–145. 26. Runyon BA, McHutchison JG, Antillon MR, et al. Short-course versus long-course antibiotic treatment of spontaneous bacterial peritonitis. Gastroenterology 1991; 100: 1737–1742. 27. Runyon BA. Low-protein-concentration ascitic fluid is predisposed to spontaneous bacterial peritonitis. Gastroenterology 1986; 91:1343–1346. 28. Andreu M, Sola R, Sitges-Serra A, et al. Risk factors for spontaneous bacterial peritonitis in cirrhotic patients with ascites. Gastroenterology 1993; 104:1133–1138. 29. Rimola A, Garcia-Tsao G, Navasa M, et al. Diagnosis, treatment and prophylaxis of spontaneous bacterial peritonitis: a consensus document. J Hepatol 2000; 32:142–153. 30. Westphal J, Jehl F, Vetter D. Pharmacological, toxicologic, and microbiological considerations in the choice of initial antibiotic therapy for serious infections in patients with cirrhosis of the liver. Clin Infect Dis 1994; 18:324–335. 31. Rimola A, Salmero´n JM, Clemente G, et al. Two different dosages of cefotaxime in the treatment of spontaneous bacterial peritonitis in cirrhosis: results of a prospective, randomized, multicenter study. Hepatology 1995; 21:674–679. 32. Franca AVC, Giordano HM, Seva-Pereira T, et al. Five days of ceftriaxone to treat spontaneous bacterial peritonitis in cirrhotic patients. J Gastroenterol 2002; 37:119–122. 33. Ariza J, Xiol X, Esteve M, et al. Aztreonam vs. cefotaxime in the treatment of gramnegative spontaneous peritonitis in cirrhotic patients. Hepatology 1991; 14:91–98. 34. Go´mez-Jimenez J, Ribera E, Gasser I, et al. Randomized trial comparing ceftriaxone with cefonicid for treatment of spontaneous bacterial peritonitis in cirrhotic patients. Antimicrob Agents Chemother 1993; 37:1587–1592.

442

Preheim

35. Grange JD, Amiot X, Grange V, et al. Amoxicillin-clavulanic acid therapy of spontaneous bacterial peritonitis: a prospective study of twenty-seven cases in cirrhotic patients. Hepatology 1990; 11:360–364. 36. Ricart E, Soriano G, Novella MT, et al. Amoxicillin-clavulanic acid versus cefotaxime in the therapy of bacterial infections in cirrhotic patients. J Hepatol 2000; 32:596–602. 37. Terg R, Cobas S, Fassio E, et al. Oral ciprofloxacin after a short course of intravenous ciprofloxacin in the treatment of spontaneous bacterial peritonitis: results of a multicenter randomized study. J Hepatol 2000; 33:564–569. 38. Navasa M, Follo A, Llovet JM, et al. Randomized, comparative study of oral ofloxacin versus intravenous cefotaxime in spontaneous bacterial peritonitis. Gastroenterology 1996; 111:1011–1107. 39. Soares-Weiser K, Brezis M, Leibovici L. Antibiotics for spontaneous bacterial peritonitis in cirrhotics. Cochrane Database Syst Rev 2001; (3): 1–16. Art. No.: CD002232. 40. Follo A, Llovet JM, Navasa M, et al. Renal impairment after spontaneous bacterial peritonitis in cirrhosis: incidence, clinical course, predictive factors, and prognosis. Hepatology 1994; 20:1495–1501. 41. Sort P, Navasa M, Arroyo V, et al. Effect of intravenous albumin on renal impairment and mortality in patients with cirrhosis and spontaneous bacterial peritonitis. N Engl J Med 1999; 341:403–409. 42. Tito´ L, Rimola A, Gine´s P, et al. Recurrence of spontaneous bacterial peritonitis in cirrhosis: frequency and predictive factors. Hepatology 1988; 8:27–31. 43. Singh N, Gayowski T, Yu VL, et al. Trimethoprim-sulfamethoxazole for the prevention of spontaneous bacterial peritonitis in cirrhosis. Ann Intern Med 1995; 122:595–598. 44. Gine´s P, Rimola A, Planas R, et al. Norfloxacin prevents spontaneous bacterial peritonitis recurrence in cirrhosis: results of a double-blind, placebo-controlled trial. Hepatology 1990; 12:716–724. 45. Grange J, Roulot D, Pelletier G, et al. Norfloxacin primary prophylaxis of bacterial infections in cirrhotic patients with ascites: a double-blind randomized trial. J Hepatol 1998; 29:430–436. 46. Ferna´ndez J, Navasa M, Go´mez J, et al. Bacterial infections in cirrhosis: epidemiological changes with invasive procedures and norfloxacin prophylaxis. Hepatology 2002; 35:140–148. 47. Campillo B, Dupeyron C, Richardet J, et al. Epidemiology of severe hospital-acquired infections in patients with liver cirrhosis: effect of long-term administration of norfloxacin. Clin Infect Dis 1998; 26:1066–1070. 48. Ortiz J, Vila MC, Soriano G, et al. Infections caused by Escherichia coli resistant to norfloxacin in hospitalized cirrhotic patients. Hepatology 1999; 29:1064–1069. 49. Rabinovitz M, Prieto M, Gavaler JS, et al. Bacteriuria in patients with cirrhosis. J Hepatol 1992; 16:73–76. 50. Kuo CH, Changchien CS, Yang CY, et al. Bacteremia in patients with cirrhosis of the liver. Liver 1991; 11:334–339. 51. Barnes PF, Arevalo C, Chan LS, et al. A prospective evaluation of bacteremic patients with chronic liver disease. Hepatology 1988; 8:1099–1103. 52. Thulstrup AM, Sørensen HT, Schønheyder HC, et al. Population-based study of the risk and short-term prognosis for bacteremia in patients with liver cirrhosis. Clin Infect Dis 2000; 31:1357–1361. 53. Selby WS, Norton ID, Pokorny CS, et al. Bacteremia and bacterascites after endoscopic sclerotherapy for bleeding esophageal varices and prevention by intravenous cefotaxime: a randomized trial. Gastrointest Endosc 1994; 40:680–684. 54. Rolando N, Gimson A, Philpott-Howard J, et al. Infectious sequelae after endoscopic sclerotherapy of oesophageal varices: role of antibiotic prophylaxis. J Hepatol 1993; 18: 290–294. 55. Kulkarni SG, Parikh SS, Dhawan PS, et al. High frequency of bacteremia with endoscopic treatment of esophageal varices in advanced cirrhosis. Indian J Gastroenterol 1999; 18:143–145.

Infections in Cirrhosis in the Critical Care Unit

443

56. Rimola A, Bory F, Teres J, et al. Oral, nonabsorbable antibiotics prevent infection in cirrhotics with gastrointestinal hemorrhage. Hepatology 1985; 5:463–467. 57. Soriano G, Guarner C, Thoma´s A, et al. Norfloxacin prevents bacterial infection in cirrhotics with gastrointestinal hemorrhage. Gastroenterology 1992; 103:1267–1272. 58. Pauwels A, Mostefa-Kara N, Debenes B, et al. Systemic antibiotic prophylaxis after gastrointestinal hemorrhage in cirrhotic patients with a high risk of infection. Hepatology 1996; 24:802–806. 59. Hsieh W, Lin H, Hwang S, et al. The effect of ciprofloxacin in the prevention of bacterial infection in patients with cirrhosis after upper gastrointestinal bleeding. Am J Gastroenterol 1998; 93:962–966. 60. Soares-Weiser K, Brezis M, Tur-Kaspa R, et al. Antibiotic prophylaxis for cirrhotic patients with gastrointestinal bleeding. Cochrane Database Syst Rev 2002; (2):1–34, Art. No. CD002907. 61. Hirota WK, Petersen K, Baron TH, et al. American Society for Gastrointestinal Endoscopy. Guidelines for antibiotic prophylaxis for GI endoscopy. Gastrointest Endosc 2003; 58:475–482. 62. Silverio RH, Perini RF, Arruda CB. Bacterial infection in cirrhotic patients and its relationship with alcohol. Am J Gastroenterol 2000; 95:1290–1293. 63. Mandell LA. Epidemiology and etiology of community-acquired pneumonia. Infect Dis Clin N Am 2004; 18:761–776. 64. Austrian R, Gold J. Pneumococcal bacteremia with especial reference to bacteremic pneumococcal pneumonia. Ann Intern Med 1964; 60:759–776. 65. Gransden WR, Eykyn SJ, Phillips I. Pneumococcal bacteremia: 325 episodes diagnosed at St. Thomas’s Hospital. Br Med J 1985; 290:505–508. 66. Chen M, Hsueh P, Lee L, et al. Severe community-acquired pneumonia due to Acinetobacter baumannii. Chest 2001; 120:1072–1077. 67. File TM, Garau J, Blasi F, et al. Guidelines for empiric antimicrobial prescribing in community-acquired pneumonia. Chest 2004; 125:1888–1901. 68. Niederman MS, Craven DE, Bonten MJ, et al. Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcare-associated pneumonia. Am J Respir Crit Care Med 2005; 171:388–416. 69. Hlady WG, Klontz KC. The epidemiology of Vibrio infections in Florida, 1981–1993. J Infect Dis 1996; 173:1176–1183. 70. Chiang S, Chuang Y. Vibrio vulnificus infection: clinical manifestations, pathogenesis, and antimicrobial therapy. J Microbiol Immunol Infect 2003; 36:81–88. 71. McCashland TM, Sorrell MF, Zetterman RK. Bacterial endocarditis in patients with chronic liver disease. Am J Gastroenterol 1994; 89:924–927. 72. Hsu, Chen RJ, Chu SH. Infective endocarditis in patients with liver cirrhosis. J Formos Med Assoc 2004; 103:355–358. 73. Gonzalez-Quintela A, Martinez-Rey C, Castroagudin JF, et al. Prevalence of liver disease in patients with Streptococcus bovis bacteremia. J Infect 2001; 42:116–119. 74. Tripodi MF, Adinolfi LE, Ragone E, et al. Streptococcus bovis endocarditis and its association with chronic liver disease: an underestimated risk factor. Clin Infect Dis 2004; 38:1394–1400. 75. Xiol X, Castellvı´ JM, Guardiola J, et al. Spontaneous bacterial empyema in cirrhotic patients: a prospective study. Hepatology 1996; 23:719–723.

22 Infections Associated with Diabetes in the Critical Care Unit Larry I. Lutwick Department of Infectious Diseases, VA New York Harbor Health Care System, and State University of New York Downstate Medical School, Brooklyn, New York, U.S.A.

INFECTION AND DIABETES As reviewed succinctly by Rajbhandari and Wilson (1), it appears to be the case that there is a measurable increased incidence of infection in those who are diagnosed with diabetes. Indeed, they cite one study of U.S. factory workers that reported that 28% of workers with diabetes took 10 or more sick-days annually related to infection as compared to 10% of controls without diabetes (2). Additionally, certain specific pathogens appear to be more prevalent in the diabetic cohort, notably Staphylococcus aureus and Candida species. The mechanism of increased infection rate in diabetics is multifactorial including direct effects of diabetes on the immune system. Among the defects found (1) have been a variety of polymorpholeukocyte function including those involving adherence, chemotaxis, and intracellular oxidative killing as well as impairment in aspects of cell-mediated immunity and monocyte function. Infection itself also produces increasing degrees of glucose intolerance. Indeed, it has been pointed out by many clinicians that a rising blood glucose is one of the earliest signs of an underlying infection. This review concentrates specifically on a number of infections that often require intensive care and in which diabetes is most commonly associated with severe disease. The topic will be divided into those recognized and reported throughout the world and reported primarily in the developed world and those reported almost exclusively in parts of the developing world. The former category will consist of emphysematous pyelonephritis (EPN), gangrenous cholecystitis, rhinocerebral mucormycosis, and Fournier’s gangrene. Malignant otitis externa, a locally invasive Pseudomonas aeruginosa of the external ear canal causing temporal bone osteomyelitis and cranial nerve involvement, is clearly a serious condition also linked to diabetes, but is not classically in need of intensive care unit care. Gangrenous cholecystitis, a more serious form of acute cholecystitis usually needing emergency cholecystectomy, is another entity that is at increased risk in the diabetic

445

446

Lutwick

patient (3). The tropical group disorders discussed are melioidosis and tropical diabetic hand syndrome (TDHS). LIFE-THREATENING INFECTION CHARACTERISTIC OF DIABETICS Emphysematous Pyelonephritis Urinary tract infections in the diabetic appear to be more common than in the nondiabetic host, though few prospective cohort observations assess this. Stapleton (4), in reviewing this topic, cited a number of observations that suggest that asymptomatic bacteriuria and symptomatic urinary tract infections are more common in the diabetic (especially the female diabetic). As an example, one study found that the incidence of urinary tract infections in postmenopausal women was twice as high in those who were diabetic (5). The same group (6) subsequently reported that during 1773 person-years of follow-up of postmenopausal women, 138 symptomatic urinary tract infections occurred (incidence, 0.07 per person-year) with diabetes being an independent predictor of infection (hazard ratio ¼ 3.4; 95% confidence interval: 1.7–7.0). Another study found that asymptomatic bacteriuria occurred four times more frequently in the diabetic (7). Complications of these lower tract infections are more common in the diabetic including Candida infections as well as emphysematous cystitis and pyelonephritis (3). EPN, a kidney infection associated with gas in and around the renal parenchyma, was first described by Kelly and MacCallum at the end of the 19th century (8) in a patient with pneumaturia. The infection is an acute necrotizing infection of the kidney itself. It is a relatively uncommon infection that continues to be associated with a high degree of morbidity and mortality. Although quite uncommon when one considers the overall number of urinary tract infections found in the diabetic host, 70% to 90% of cases of this entity occur in the diabetic (9). Most diabetic patient with EPN, but not all, have issues with adequate glucose control but not necessarily manifesting ketoacidosis (10). In those individuals without diabetes, the most common comorbidity is obstructive uropathy, but polycystic kidney disease, end stage nephropathies, and immunosuppression have been linked to it (11). Obstructive uropathy is generally found in a great majority of nondiabetics who develop EPN, and 50% of diabetics (11). EPN, being more common in women, reflects the increased number of urinary tract infections in women as compared to men, diabetic or not. The left kidney appears to be the more common side affected (60%), and bilateral involvement occurs in the remaining 5% of cases (10). Its clinical presentation is similar to that of the more common acute pyelonephritis (nonemphysematous) with fever and chills associated with flank, costovertebral angle, and/or abdominal pain. More prominent symptoms can suggest EPN such as lethargy, confusion, low platelet count, increasing azotemia, and overt shock, all symptoms suggesting an ongoing sepsis syndrome. One uncommon but helpful physical finding is the presence of crepitation over the patient’s flank with or without a clearly palpable mass (12). The EPN is diagnosed radiographically (13) by visualizing gas in the renal parenchyma and the perinephric space. The gas is thought to be produced by the fermentation of the high tissue concentrations of glucose by the etiologic microorganisms. The gas can dissect further the subcapsular and perinephric spaces and can be found in the contralateral retroperitoneal space whether the other kidney is affected or not (11). It may even dissect along the psoas margin into the scrotal sac and the spermatic cord. The standard abdominal radiogram can reveal mottled

Infections Associated with Diabetes in the Critical Care Unit

447

collections of gas in and around the renal parenchyma. Radiographic staging has been proposed by Michaeli et al. (14): Stage 1: Gas in either the renal tissue or the pericapsular areas Stage 2: Gas in both the parenchyma and the pericapsular areas Stage 3: Extension through Gerota’s fascia (the renal capsule) and/or bilateral disease Nephrolithiasis may also be visualized, which may have been a significant comorbidity in the development of the EPN. Ultrasound and/or computed tomography (CT) may also reveal the gas patterns. Ultrasound is not as good as CT for elucidating the gas patterns and is a more operator-dependent procedure. CT is generally considered to be the diagnostic test of choice because it can well define the extent and amount of gas, can assess destruction of the renal tissue by the process, and can be useful in assisting the placement of one or more catheters for drainage. The destruction of the renal parenchyma may in part be due to infection per se but swelling of the kidney in its capsule, which may impair blood supply and/or renal vessel thrombosis, may play a role. Additionally, CT is an excellent technique for following the response to therapy as carbon dioxide is usually rapidly absorbed and prolonged persistence implies poor response (13). Gas may be found entirely in the collecting system of a kidney, a condition referred to as emphysematous pyelitis. It can present similarly, albeit often less severely, is usually associated with diabetes as well, and generally has a lower degree of morbidity and mortality, but is not inconsequential, however. Gas may also be found entirely in the urinary bladder, so called emphysematous cystitis. This entity, which also can be linked with the diabetic host, is much more often associated with pneumaturia and can also be linked to fistulae from the colon or vagina communicating with the urinary bladder. This can be associated with either a malignant or nonmalignant process (i.e., diverticulitis). The microbiology of EPN is quickly obvious on culture, but antimicrobial therapy should not be delayed awaiting culture results. The process is overwhelming due to the facultative enteric gram-negative bacilli, the most common of which in most studies is Escherichia coli, the most common cause of urinary tract infections overall. Other gram-negative bacilli to be commonly linked to EPN are Proteus mirabilis and Enterobacter aerogenes. In Shokeir’s report (10), of the 15 patients who had blood cultures done, all of them grew the same bacterium (or bacteria as mixed infections may occur) that was also isolated from urine culture. Occasionally other gram-negative bacilli such as P. aeruginosa (a bacterium inherently more resistant to antimicrobials than some of the others) may be isolated, more often in cases associated with recurrent urinary tract infections in the past and with multiple courses of antimicrobial agents. Rarely, diverse organisms such as the yeasts, Candida albicans or Cryptococcus neoformans, anaerobic streptococci, and the phylogenetically confused Pneumocystis carinii have been reported to be associated with EPN (14). The therapy needed for EPN has generally been considered to be active antimicrobial therapy, glucose control, and nephrectomy (10). Even with these modalities, an overall mortality rate of 30% to 40% may be found. In some less severe cases, however, medical therapy alone or combined with percutaneous drainage has been successful (15–17). The initial choice of empirical antimicrobial agents, prior to culture results, can be quite vital in assisting in a favorable outcome. Rational choice requires the following: 1. Considering the common resistance pattern of gram-negative bacilli to antimicrobials in the geographic area of the patient. As an example, the

448

Lutwick

sensitivity to a fluoroquinolone may be much lower in New York City than in Kalamazoo. 2. Knowing the patient’s history of antimicrobial sensitivity. If it is reported that a patient is allergic to an antimicrobial that might be considered for use then it is important to assess what the manifestation of allergy is. For example, nausea and vomiting after receiving an oral antimicrobial does not preclude its use especially parenterally. Likewise, a nonspecific drug–associated rash (maculopapular nonurticarial eruption) does not predict serious IgE-associated anaphylaxis if used again. Indeed, the incidence of anaphylaxis following no previous reaction to penicillin use is not different from after a nonspecific rash from penicillin. 3. Being cognizant of the so-called banana peel syndrome. This euphemism suggests broader initial antimicrobial therapy for individuals who are seriously ill, ‘‘one foot in the grave and the other on a banana peel.’’ This does not at all preclude the narrowing of the spectrum of the therapy once the antimicrobial resistance pattern of the pathogen is known. Rhinocerebral Zygomycosis Rhinocerebral zygomycosis (RCZ, also referred to as rhinocerebral mucormycosis) is an acute and often fatal fungal infection of the nasal mucosa and adjacent cerebral parenchymal and vascular tissues. Although this infection can occur in seemingly healthy individuals, RCZ is, in general, linked to diabetics. The diabetic state is classically one with ketoacidosis. As an example, in a Mexican series of 22 cases of RCZ (of 36 cases of zygomycosis overall), 20 were in diabetics (18). In the diabetic cases, 10 had ketoacidosis, one had hyperosmolar coma, and nine were ‘‘stable.’’ Other reviews report that between 60% and 80% of cases of RCZ had diabetes, with half of these with ketoacidosis. These organisms are associated with a ketone reductase system that facilitates the growth of the molds in high glucose, acidotic, and ketotic milieu (19). The other two cases in the Mexican series had myelodysplasia and chronic renal failure as cofactors. Indeed, acute leukemia is associated with RCZ as well, and additional cases have been linked to severe malnutrition, steroid therapy, desferrioxamine toxicity, and severe burns. The demographics of other forms of zygomycosis (pulmonary, cutaneous, and disseminated) are much less likely to be associated with the diabetic state. Rhizopus species are the most commonly isolated agents of this severe infection, followed by Absidia, Rhizomucor and Cunninghamella. Laboratory confirmation of the identity of the organism is the only way to differentiate among the fungi. They are ubiquitous fungi that are common inhabitants of decaying matter. As an example, Rhizopus spp. can be recovered frequently from moldy bread. The spores of these fungi causing RCZ gain entry to the body through the respiratory tract and presumably are deposited on the nasal turbinates. Normal human serum can inhibit the growth of Rhizopus. In contrast, serum obtained from patients with diabetic ketoacidosis is not inhibitory and may actually enhance fungal growth (20). Once the ubiquitous fungus begins to grow, the wide, nonseptate, right angle branching hyphae invade tissue and have a special affinity for blood vessels. Direct penetration and growth through the arterial blood vessel wall explain the propensity for thrombosis and tissue necrosis, major hallmarks of the pathology of this infection. Lymphatic vessels and nerves can also be directly invaded. Progressive infection dissects the internal elastic lamina from media of the artery, leading to extensive endothelial damage and thrombosis, causing infarction in the tissues supplied.

Infections Associated with Diabetes in the Critical Care Unit

449

It appears clear that early recognition of the initial signs and symptoms of RCZ are quite important in management. The infection begins at the nasal mucosa and spreads quickly (within days) to the adjacent paranasal sinuses, the orbit, and via direct extension to the brain through the ethmoid bone or along penetrating vessels. The method of penetration into the brain may also be via the cribriform plate, which is thin and also has preformed pathways in the form of olfactory nerves passing through it, the roof of orbit, which is also very thin, or through the retro-orbital region. Cavernous sinus thrombosis and/or carotid artery involvement may occur. Initial symptoms of RCZ include some degree of fever, altered vision, and facial swelling. In a patient with diabetic ketoacidosis, an alert diagnostician should suspect RCZ when the altered mental status of diabetic ketoacidosis does not improve within a day or so of correction of the metabolic abnormalities. Other symptoms include facial pain, headache, and nasal stuffiness. In a large review of 114 cases (21), no symptom was reported in more than 44% of affected individuals. On physical examination, invasion of the nasal mucosa with associated tissue infarction can produce a necrotic, black eschar, which may be visible. Early on, these lesions are visible in about 20% of patients, but close to 40% will develop them at some time (22). More lesions will be seen with the use of endoscopic examination as compared to routine rhinoscopy. Dark necrotic epistaxis may be the only visual finding (22). It is important to note that biopsy of this area may not demonstrate the organism, only infarction, because the fungi are generally found deeper in the tissue. Other physical signs of RCZ are facial edema and multiple cranial nerve palsies, as the infection spreads into the orbital apex. The orbital apex syndrome is associated with unilateral ptosis, proptosis, visual loss, complete ophthalmoplegia, maxillary and ophthalmic nerve anesthesia, and anhidrosis (23). Plain roentgenograms of the sinuses and orbits can reveal sinusoidal mucosal thickening, with or without air-fluid levels. Erosion of bone through the walls of the sinuses or into the orbit can be found as the disease progresses. Destruction of bone in this region is often dramatically revealed by CT. Abnormalities in soft tissues involved in the disease process can also be visualized by CT scans and can be used to guide surgical intervention. If disease is seen, nasal endoscopy with biopsy from tissue is mandatory. The specimen should be sent for fungal staining (a 10% KOH mount may reveal the hyphal elements), fungal culture, and histopathology. Fixed tissue can be stained with hematoxylin and eosin, and fungal hyphae can be seen with this routine histologic stain. Grocott methenamine-silver or periodic acid-Schiff staining also adequately demarcates fungal elements in tissue in most cases. Despite the availability of a variety of new azole antifungal medications (such as itraconazole and voriconazole) and echinocandins (caspofungin and micafungin), the standard drug for RCZ has remained to be amphotericin B (24). Because the agents of zygomycosis are relatively refractory to medical treatment, the maximum tolerated dose of the drug is used, typically 1.0 to 1.5 mg/kg/day. This dosage range is usually associated with renal function abnormalities, which can limit the use of the drug. There are some successful outcomes in patients with RCZ when treated with lipid preparations of amphotericin B (25). In RCZ, the recommended dose of lipid formulations of amphotericin B is 5 mg/kg daily. A new, broad-spectrum triazole, posaconazole (not yet commercially available in the United States), has been shown to be active in a murine model of zygomycosis (26). Although patients recover from RCZ using antifungal therapy alone, these are clearly the exception, and aggressive surgical de´bridement of necrotic tissue is

450

Lutwick

advisable. As reviewed by Sugar (24), some patients may recover with minimally disfiguring surgery. A medical-surgical approach, however, improves the chances of success. Repeated operations may be required for satisfactory removal of continuously appearing necrotic tissue. Major reconstructive surgery may also be necessary during convalescence if the patient survives. Although hyperbaric oxygen therapy has been utilized, this therapy is not routinely recommended at present (24). Two factors determine the outcome in all patients: early diagnosis and resolution of predisposing problems (24). The overall mortality rate has been about 50% (24), although higher survival rates, up to 80% (2), have been reported more recently. This compares with survival rates of about 12% in 1961 (27). Yohai et al. (21) reported that treatment within six days of symptom onset produced a survival rate of about 80%, whereas delay for more than 12 days after onset of symptoms resulted in a 40% survival rate. Diabetics (where the underlying hyperglycemia and ketosis can be reversed) had a 77% survival as compared with a rate of 34% in nondiabetics (21). Similar numbers have been reported in other studies (28).

Fournier’s Gangrene Necrotizing fasciitis was characterized by the Confederate Army surgeon Joseph Jones in the postwar period of 1871. When the process, a necrotizing, gas-producing infection spreading quickly along the fascial planes, involves the perineum (and scrotum in the male patient), it is referred to as Fournier’s gangrene. It was first described by Alfred Jean Fournier, a Parisian venereologist, in 1843 (29). Although single organisms (such as the Group A beta-hemolytic streptococcus) can be associated with necrotizing fasciitis, Fournier’s gangrene is clearly a polymicrobial process including gram-positive cocci such as streptococci and facultative gram negative bacilli such as E. coli and Proteus as well as a variety of strict anaerobic bacteria including Bacteroides, Peptostreptococcus, and Fusobacterium. The process has been historically associated with diabetes among other risk factors. Other factors that may predispose to this progressive, destructive process include alcoholism, local pathology (such as rectal abscesses, hidradenitis, and urinary tract infections with strictures), blunt local trauma, postsurgical complications, and the injection of illicit drugs into the superficial penile veins (30). In one relatively recent report, diabetes was second only to perianal pathology as a predisposing cause of Fournier’s gangrene (31). An uncommon problem leading to intensive care for antimicrobial and surgical interventions has been estimated to occur in 1 in 7500 hospital admissions (32) or 1% of urological admissions in another (33). Usually initially beginning as pain and/or pruritus in or around the scrotum with fever and chills, the process quickly manifests as a cellulitic area in the scrotum or peritoneum with very prominent pain and marked systemic toxicity. This progresses quickly to prominent soft tissue swelling of the genitalia usually with subcutaneous crepitus. These changes can spread superiorly to the anterior abdominal wall, inferiorly to the anterior thighs and posteriorly to the perianal areas (33). Dark purple patches develop in the area and progress to extensive scrotal necrosis. At this point, local pain dramatically decreases, probably related to destruction of the sensory nerves. If it is not treated immediately, gangrenous sloughing of the tissue will ensue. Radiographically, the detection of gas in the scrotal and perineal tissues increases from the yield by palpation of crepitus on examination (64%) (34) to 90%

Infections Associated with Diabetes in the Critical Care Unit

451

on the standard X ray (30). The pattern of gas is sometimes referred to as a ‘‘honeycomb’’ scrotum. An ultrasonic examination of the scrotum (30) demonstrates prominent thickening of the scrotal skin with subcutaneous gas. Scrotal hernias can demonstrate gas on the scrotal ultrasound, but it is the gas in the hernial sac, not in the scrotal skin. Additional findings can be peritesticular fluid collections. Treatment must include aggressive broad-spectrum antimicrobial therapy aimed at gram-positive cocci, facultative enteric gram negative, and strict anaerobes. The foul odor of the process supports a prominent role of anaerobes in the necrotic infection. Usual regimens include those aimed at usual bowel flora (antipseudomonal penicillins, carbapenems, or a combination such as ampicillin/sulbactam plus an aminoglycoside, aztreonam, or fluoroquinolone). Surgical intervention usually necessitates debridement of the scrotal sac and, often, bilateral orchiectomy. As in many necrotic deep tissue infections, after debridement extensive restorative and reconstructive surgery may be needed. LIFE-THREATENING INFECTION CHARACTERISTIC OF DIABETICS IN THE TROPICS Most reviews regarding critical care for infections in diabetics do not focus or even mention infections that are exclusively (or almost so) described in the tropics. Two of these will be discussed here, melioidosis and TDHS. Neither of these infections is well known in the developed world. Melioidosis is particularly relevant because it is considered to be a class B bioterrorism infection. Melioidosis The causal organism, Burkholderia pseudomallei, has, as many before it, gone through many name changes from Loefflerella or Pfeifferella whitmori and Bacillus or Pseudomonas pseudomallei to, in 1992, its current designation. The genus is named after Walter Burkholder who first characterized Burkholderia cepacia as a phytopathogen responsible for a root rot of onions. B. pseudomallei is a motile, aerobic, and nonspore-forming gram-negative bacillus. Although primarily an intracellular organism, it readily grows on most solid media resulting in prominently wrinkled (rugose) colonies that may manifest an earthy-like aroma. Selective media are available for isolation as well. Gram stain of the bacillus can reveal the safety pin bipolar appearance often seen with Yersinia pestis. There are some B. pseudomallei–like organisms that are much less virulent. Formerly considered to be a separate biotype, these L-arabinoside assimilators are now classified as Burkholderia thailandensis and account for about a quarter of soil isolates in Thailand (35). B. pseudomallei is a hard-core survivalist organism nutritionally versatile to persist in triple-distilled water for long periods of time (36). The organism exists in nature as an environmental saprophyte that lives in the soil and surface water in endemic areas (Southeast Asia and northern, tropical Australia), particularly in rice paddies (37–39). In endemic countries, the organism exists primarily in focal areas and is not equally distributed throughout the landscape. Sporadic cases have been reported to be acquired in parts of Africa and the Americas. Two recent outbreaks in Australia have also implicated potable water supplies rather than surface water as a potential source of the infection (40,41). Melioidosis is a disease of rainy season in these endemic areas (37,42). It mainly affects people who have direct contact with soil and water. Many have an underlying

452

Lutwick

predisposing condition such as diabetes (which overall is the most common risk factor), renal disease, cirrhosis, thalassemia, alcohol dependence, immunosuppressive therapy, chronic obstructive lung disease, and cystic fibrosis (42). Melioidosis may present at any age, but peaks in the fourth and fifth decades of life, affecting men more than women. In addition, although severe fulminating infection can and does occur in healthy individuals, severe disease and fatalities are much less common in those without risk factors. Infection in humans is usually acquired by inoculation in an open wound or inhalation of aerosolized soil or water and not generally by ingestion. Although inhalation of aerosolized organisms causing pneumonia clearly occurs, pneumonia has also occurred following well-documented skin injuries (43), suggesting that the lung involvement can be related to bacteremic spread as well. The incubation period after significant exposure to B. pseudomallei can be as short as one day but averages about nine days; however, because of ‘‘latency’’ (the mechanism of which is unclear) it can be up to 29 years. Recrudescent infections in veterans of the Vietnam War have given rise to the nickname ‘‘Vietnamese Time Bomb’’ (44). Despite the risk of reactivation, documented American cases were fairly uncommon as compared to the individuals exposed in Vietnam. An Australian study, in fact, suggested that only 3% of melioidosis infections were related to reactivation, and 97% to acute disease (45). Melioidosis presents mostly as a febrile illness, ranging from an acute fulminant septicemia to a chronic debilitating localized infection to an unknown subclinical infection. As virtually every organ can be affected, melioidosis has been termed a ‘‘great imitator’’ of many other infectious diseases (46). The majority of infected patients are asymptomatic. The most commonly recognized presentation of melioidosis is pneumonia with high fever, myalgias, and chest pain. Although the cough can be nonproductive, respiratory secretions may be purulent, significant in quantity, and associated with intermittent hemoptysis. The process can be rapidly fatal with bacteremia and hypotension. In addition to an acute pneumonia which may result in intensive care unit admission, chronic pulmonary infection may also be caused by B. pseudomallei, either as a continuum for acute disease or as reactivation years later. The presentation is quite similar to reactivation tuberculosis with upper lobe involvement associated with productive cough, weight loss, and hemoptysis. Acute melioidosis septicemia is the most severe complication of the infection. It presents as a typical sepsis syndrome with hypotension, high cardiac output, and low systemic vascular resistance. In many cases, a primary focus in the soft tissues or lung can be found. The syndrome, usually in patients with risk factor comorbidities, is characteristically associated with multiple abscesses involving the cutaneous tissues, the lung, the liver and spleen, and a very high mortality rate of 80% to 95%. With prompt optimal therapy, the case fatality rate can be decreased to 40% to 50%. In acute severe melioidosis, there is the rapid progression of respiratory failure that is due to acute respiratory distress syndrome and/or pneumonia. It has been suggested that the acute respiratory distress syndrome (ARDS) to melioidosis sepsis is more rapid in progression than with other bacteria and may be related to the intracellular interactions of the bacillus and the leukocyte (47). Bacteremia without shock/hypotension has a substantially better prognosis. Abscesses can be found in many organs. Two organs that are particularly relevant in disease are the prostate and the parotid gland. Acute prostatic abscess may

Infections Associated with Diabetes in the Critical Care Unit

453

cause urinary retention. Residual prostatic abscess appears to be a potential focus for reactivation infection or relapse and unlike other visceral collection of melioidosis, unless the abscesses are large and accessible, ought to be definitively drained as needed. The purulent material obtained is yellow to tan in color and odorless. In focal melioidosis without bacteremia, the mortality rate is 4% to 5%. The melioidosis bacillus is intrinsically insensitive to many antimicrobials. B. pseudomallei is usually inhibited by tetracyclines, chloramphenicol, trimethoprimsulfamethoxazole (SXT), antipseudomonal penicillins, carbapenems, ceftazidime, and amoxicillin/clavulanate or ampicillin/sulbactam. Ceftriaxone and cefotaxime have good in vitro activity but poor efficacy (M35), and cefepime does not appear to be equivalent to ceftazidime in a mouse model (48). Samuel and Ti (49) have reviewed the randomized and quasirandomized trials comparing melioidosis treatment and found that the formerly standard therapy of chloramphenicol, doxycycline, and SXT combination had a higher mortality rate than therapy with ceftazidime, imipenem/cilastatin, or amoxicillin/clavulanate (or ampicillin/sulbactam). The betalactam-betalactamase inhibitor therapy, however, seemed to have a higher failure rate. A more prolonged oral phase of treatment is used to decrease the risk of late relapse with a total period of therapy of 20 weeks. During the oral therapy phase, the conventional standard regimen appears to be equivalent to any newer therapies. Table 1 lists current treatment recommendations (50).

Tropical Diabetic Hand Syndrome (TDHS) In the developed world, diabetic infections of the lower extremity remain a significant cause of morbidity leading to disability, prolonged hospital stays, and amputations of Table 1 Treatment of Burkholderia pseudomallei Infectiona Initial parenteral therapy for severe infection (usual 14 day minimum) Ceftazidimeb: 40 mg/kg IV every 8 hrs (typical adult dose 2 g) or Imipenem/cilastatinc: 20 mg/kg IV every 6–8 hrs (typical adult dose 1 g) (note: IV amoxicillin/clavulanate or ampicillin/sulbactam can be used in every 4 hrs dosing but is associated with a higher failure rate) Follow-up oral therapy (to complete 20 wks of treatment) (note: in mild, localized disease, oral therapy can be used for the entire 20 wks) Doxycycline: 2 mg/kg orally (PO) every 12 hrs (typical adult dose 100–200 mg) and Trimethoprim-sulfamethoxazole (fixed 1:5 combination): typical adult dose two double strength (trimethoprim 320/sulfamethoxazole 1600) PO every 12 hrs and Chloramphenicol: 10 mg/kg PO every 6 hrs for the first 8 wks (typical adult dose 500–1000 mg) or (especially in children or pregnant women) Amoxicillin/clavulanate (fixed combination 2:1): 10 mg/kg amoxicillin/5 mg/kg clavulanate PO every 8 hrs (typical adult dose 1000 mg/500 mg) and Amoxicillin: 10 mg/kg PO every 8 hrs (typical adult dose 1000 mg) a

Dosing may require adjustments in renal or hepatic dysfunction. Ceftriaxone and cefotaxime have good in vitro activity but a higher mortality rate and should not be used. No human data is found for cefepime. c Meropenem, 1 g or 25 mg/kg IV every eight hours, may be used in lieu of imipenem/cilastatin. b

454

Lutwick

toes, feet, and sometimes lower legs. It is not common, however, that such infections result in intensive care unit stays. Both in the developed and developing world, the most relevant risk factor in the diabetic appears to be an underlying peripheral neuropathy. This denervation of the sensory nerves of the foot impairs the perception of traumatic events (including ill-fitting shoes), resulting in the development of calluses, cracking soles, fissures, and other direct breakdown of the protective skin to produce ulcerations and infection. When combined with large and small vessel peripheral vascular disease and changes in polymorphonuclear leukocyte function associated with hyperglycemia, a range of infection may develop from cellulitis to deeper soft tissue infections to osteomyelitis related to deeper contiguous spread of infection (51). In parts of the tropical world, however, a similar but even more aggressive infection, requiring intensive care unit, has been recognized. This severe and limband even life-threatening upper extremity sepsis is TDHS. This condition is far less recognized in the developed world but is a significant cause of both morbidity as well as mortality in parts of the African continent (52,53) and has been described in India as well (54). Although a report of a similar syndrome was initially described in the United States (55), the African experience was first published in 1984 from Nigeria (56). In this study, 3% of 152 consecutive hospitalized diabetics were found to develop ulcerations of the hand and frank gangrene of the extremity. All of the five patients had progressive disease associated with initial trivial hand trauma. Importantly, none of the individuals had clinical evidence of either peripheral neuropathy or peripheral vascular disease. As reviewed by Abbas et al. who have published extensively regarding this entity, cases have been reported from a number of areas of the African continent including Tanzania, Kenya, and Libya (53). Many of the earlier studies were primarily descriptive so that risk factors were difficult to clearly elucidate, but proposed risk factors included insect bites or other minimal hand trauma, adult-onset diabetes, female gender, delayed seeking of medical care, poor glucose control, low socioeconomic status, and living near a coastal area. In 2001, a case–control study of TDHS was reported from Das-es-Salaam, Tanzania (a coastal city) involving 31 patients and 96 control diabetics (52). Despite the previously postulated risk factors, this study found that on logistic regression analysis independent risk factors were low body mass index, type 1 diabetes, and peripheral neuropathy. The initial wounds reflected the spectrum of hand injuries including insect bites, burns, other traumatic injuries, and nonspecific papules. At the time of presentation of TDHS, more than 80% had had purulent hand ulcerations and almost 30% rapidly progressed to frank gangrene. Thirteen percent of the total needed arm amputation for progressive gangrene despite glucose control and antimicrobial therapy, and another 13% died from unbridled sepsis despite aggressive therapy. In a subsequent report expanding the numerator of involved cases published in 2002 (57), 72 individuals fitting the case definition were included. The case definition was unchanged from previous reports and was any adult diabetic (greater than 18 years old) who had sought medical attention with cellulitis or other deeper infection of the hand with or without gangrene. Here 61% had type 2 diabetes, with an average age of 52 years (range 20–89), median interval of five years since the diagnosis of diabetes (two weeks to 19 years), low median body mass index, and generally high glucose levels. Of note, only 10 (14%) had evidence of peripheral neuropathy. In this study (57), at the time of presentation, each of the affected individuals had ulcerations with 85 of them purulent in nature, 32% had deep ulcerations involving bone, and a quarter already had localized or progressive gangrenous changes. The median

Infections Associated with Diabetes in the Critical Care Unit

455

time to presentation from perceived onset of symptoms was two weeks but was as short as two days and as long as 252 days. In most of the reports regarding TDHS, little, if any, microbiological information is given, which could reflect the lack of adequate microbiological support. The Centers for Disease Control and Prevention report, from the same Tanzanian authors, however, notes that superficial swab cultures all revealed polymicrobial growth reflective of the pathogens often found in Meleney’s synergistic gangrene. These organisms included gram-positive cocci such as staphylococci and streptococci and gram-negative bacilli such as Klebsiella pneumoniae, E. coli, and P. aeruginosa. No anaerobic flora were noted but were likely to be present. It was felt that these superficial swabs were inadequate to guide the choice of antimicrobial agents and advised tissue biopsy cultures in this regard. In this study, half of the individuals required surgery, and of this, 44% had gangrene, and about half of these needed amputation of digits, hand or arm. More than half of the group able to be followed up had enough impaired hand function to affect their activities of daily living, and many reported severe, ongoing neuropathic pain. REFERENCES 1. Rajbhandari SM, Wilson RM. Unusual infections in diabetes. Diabetes Res Clin Pract 1998; 39:123–128. 2. Wilson RM. Pickup J, Williams G, eds. Textbook of Diabetes. 2. Oxford: Blackwell, 1991: 813–831. 3. Fagan SP, Awad SS, Rahwan K, et al. Prognostic factors for the development of gangrenous cholecystitis. Am J Surg 2003; 186:481–485. 4. Stapleton A. Urinary tract infections in patients with diabetes. Am J Med 2002; 113(1A): 80S–84S. 5. Boyko E, Fihn S, Scholes D, et al. Diabetes mellitus and the risk of acute urinary tract infection among post-menopausal women. Diabetes Care 2002; 25:1778–1783. 6. Jackson SL, Boyko EJ, Scholes D, et al. Predictors of urinary tract infection after menopause: a prospective study. Am J Med 2004; 117:903–911. 7. Geerlings SE, Stolk RP, Camps MJ, et al. Asymptomatic bacteriuria may be considered a complication in women with diabetes. Diabetes Care 2000; 23:744–749. 8. Kelly HA, MacCallum WG. Pneumaturia. JAMA 1898; 31:375. 9. McDermid, Watterson J, van Eeden SF. Emphysematous pyelonephritis: case report and review of the literature. Diabetes Res Clin Pract 1999; 44:71–75. 10. Skokeir AA, El-Azav M, Mohsen T, El-Diasty T. Emphysematous pyelonephritis: a 15-year experience with 20 cases. Urology 1997; 49:343–346. 11. Stone SC, Mallon WK, Childs JM, Docherty SD. Emphysematous pyelonephritis: clues to rapid diagnosis in the emergency department. J Emerg Med 2005; 28:315–319. 12. Bonoan JT, Mehra S, Cunha BA. Emphysematous pyelonephritis. Heart Lung 1997; 26: 501–503. 13. Narlawar RS, Raut AA, Nagar A, et al. Imaging features and guided drainage in emphysematous pyelonephritis: a study of 11 cases. Clin Radiol 2004; 59:192–197. 14. Michaeli J, Mogle MJ, Heiman PS, Cains HS. Emphysematous pyelonephritis. J Urol 1984; 131:203–207. 15. Najjar M, Gouda HE, Rodriguez P, Ahmed S. Successful medical management of emphysematous pyelonephritis. Am J Med 2002; 113:262–263. 16. Chen MT, Huang CN, Chou YH. Percutaneous drainage in the treatment of emphysematous pyelonephritis: a 10 year experience. J Urol 1997; 157:1569–1573. 17. Cardinael AS, De Blay V, Gilbeau JP. Emphysematous pyelonephritis: successful treatment with percutaneous drainage. Am J Roentgenol 1995; 164:1554–1555.

456

Lutwick

18. Rangel-Guerra RA, Martinez HR, Saenz C, et al. Rhinocerebral and systemic mucormycosis. Clinical experience with 36 cases. J Neurol Sci 1996; 143:19–30. 19. Anand VK, Alemar G, Griswold JA. Intracranial complications of Mucormycosis: an experimental model and clinical review. Laryngoscope 1992; 102:656–662. 20. Gale GR, Welch A. Studies of opportunistic fungi: I. Inhibition of R. oryzae by human sera. Am J Med Sci 1961; 45:604–612. 21. Yohai RA, Bullock JD, Aziz AA, Mardert RJ. Survival factors in rhino-orbital-cerebral mucormycosis. Surv Ophthalmol 1994; 39:3–22. 22. Hendrickson RG, Olshaker J, Duckett O. Rhinocerebral mucormycosis: a case of a rare, but deadly disease. J Emerg Med 1999; 17:641–645. 23. Schwartz JC. Rhinocerebral mucormycosis. Three case reports and subject review. J Emerg Med 1985; 3:11–19. 24. Sugar AM. Agents of mucormycosis and related species. In: Mandell GL, Bennett JE, Dolin R, eds. Mandell, Douglas and Bennett’s: Principles and Practice of Infectious Diseases. Philadelphia, PA: Elsevier, Churchill, Livingstone, 2005:2973–2984. 25. Strasser MD, Kennedy RJ, Adam RD. Rhinocerebral mucormycosis: therapy with amphotericin B lipid complex. Arch Intern Med 1996; 156:337–339. 26. Sun QN, Najvar LK, Bocanegra R, et al. In vivo activity of posaconazole against Mucor spp. in an immunosuppressed mouse model. Antimicrob Agents Chemother 2003; 46: 2310–2312. 27. Ferry AP. Cerebral mucormycosis (phycomycosis). Ocular findings and review of the literature. Surv Ophthalmol 1961; 6:1–24. 28. Blitzer A, Lawson W, Meyers BR, Biller HF. Patient survival factors in paranasal sinus mucormycosis. Laryngoscope 1980; 90:635–648. 29. Efem SE. The features and aetiology of Fournier’s gangrene. Postgrad Med J 1994; 70: 568–571. 30. Morrison D, Blaivas M, Lyon M. Emergency diagnosis of Fournier’s gangrene with bedside ultrasound. Am J Emerg Med 2005; 23:544–547. 31. Capitan Manion C, Tejido Sanchez A, Suarez Charneco A, et al. Fournier’s gangrene. A serious infectious disease. Eur Urol Suppl 2003; 2:18. 32. Bejanga BI. Fournier’s gangrene. Br J Urol 1979; 51:312–316. 33. Hejase MJ, Simonin JE, Bihrle R, Coogan CL. Genital fournier’s gangrene: experience with 38 patients. Urology 1996; 47:734–739. 34. Smith MD, Angus BJ, Wuthiekanun V, White NJ. Arabinose assimilation defines a nonvirulent biotype of Burkholderia pseudomallei. Infect Immun 1997; 65:4319–4321. 35. White NJ. Melioidosis. Lancet 2003; 361:1715–1722. 36. Chaowagul W, White NJ, Dance DA, et al. Melioidosis: a major cause of communityacquired septicemia in northeastern Thailand. J Infect Dis 1989; 159:890–899. 37. Wuthiekanun V, Smith MD, Dance DAB, White NJ. The isolation of Pseudomonas pseudomallei from soil in northeastern Thailand. Trans R Soc Trop Med Hyg 1995; 89:41–43. 38. Strauss JM, Groves MG, Mariappan M, et al. Melioidosis in Malaysia. II. Distribution of Pseudomonas pseudomallei in soil and surface water. Am J Trop Med Hyg 1969; 18: 698–702. 39. Currie BJ, Fisher DA, Howard DM, et al. Endemic melioidosis in tropical northern Australia: a 10-year prospective study and review of the literature. Clin Infect Dis 2000; 31:981–986. 40. Currie BJ, Mayo M, Anstey NM, et al. A cluster of melioidosis cases from an endemic region is clonal and is linked to the water supply using molecular typing of Burkholderia pseudomallei isolates. Am J Trop Med Hyg 2001; 65:177–179. 41. Leelarasamee A, Bovornkitti S. Melioidosis: review and update. Rev Infect Dis 1989; 11: 413–425. 42. Suputtamongkol Y, Chaowagul W, Chetchotisakd P, et al. Risk factors for melioidosis and bacteremic melioidosis. Clin Infect Dis 1999; 29:408–413.

Infections Associated with Diabetes in the Critical Care Unit

457

43. Currie BJ, Fisher DA, Howard DM, et al. The epidemiology of melioidosis in Australia and Papua New Guinea. Acta Trop 2000; 74:121–127. 44. Goshorn RK. Recrudescent pulmonary melioidosis. A case report involving the so-called ‘‘Vietnamese Time Bomb.’’ Indiana Med 1987; 80:247–249. 45. Currie BJ, Fisher DA, Anstey NM, Jacups SP. Melioidosis: acute and chronic disease relapse and reactivation. Trans R Soc Trop Med Hyg 2000; 94:301–304. 46. Poe RH, Vassalo CL, Domm BM. Melioidosis: the remarkable imitator. Am Rev Respir Dis 1971; 104:427–431. 47. Puthucheary SD, Vadivelu J, Wong KT, Ong GSY. Acute respiratory failure in melioidosis. Singapore Med J 2001; 42:117–121. 48. Ulett GC, Hirst R, Bowden B, et al. A comparison of antibiotic regimens in the treatment of acute melioidosis in a mouse model. J Antimicrob Chemother 2003; 51:77–81. 49. Samuel M, Ti TY. Interventions for treating melioidosis (Cochrane Review). In: The Cochrane Library, Issue 4. Chichester, U.K.: John Wiley & Sons, Ltd., 2003. 50. Tolaney P, Lutwick LI. Melioidosis. In: Lutwick LI, Lutwick SM, eds. Bioterror: The Weaponization of Infectious Diseases. Towana, NJ: Humana Press. In press. 51. Lipsky BA. A current approach to diabetic foot infections. Curr Infect Dis Rep 1999; 1:253–260. 52. Abba ZG, Lutale J, Gill VG, Archibald LK. Tropical diabetic hand syndrome: risk factors in an adult diabetes population. Int J Infect Dis 2001; 5:19–23. 53. Abbas ZG, Gill GV, Archibald LK. The epidemiology of diabetic limb sepsis: an African perspective. Diabetic Med 2002; 19:895–899. 54. Bajaj S, Bajaj AK. Tropical diabetic hand syndrome—Indian experience. J Assoc Physicians India 1999; 47:1118–1119. 55. Mann RJ, Peacock M. Hand infections in patients with diabetes mellitus. J Trauma 1977; 17:376–380. 56. Akintewe TA. The diabetic hand—5 illustrative case reports. Br J Clin Pract 1984; 38: 368–371. 57. Centers for Disease Control and Prevention. Tropical diabetic hand syndrome-Dar es Salaam, Tanzania, 1998-2002. Morbid Mortal Wkly Rep 2002; 51:969–970.

23 Infection in Organ Transplant Patients in the Critical Care Unit Patricia Mun˜oz Clinical Microbiology and Infectious Diseases Department, Hospital General Universitario ‘‘Gregorio Maran˜o´n,’’ Universidad Complutense, Madrid, Spain

Almudena Burillo Department of Clinical Microbiology, Hospital Madrid-Monteprı´ncipe, Madrid, Spain

Emilio Bouza Clinical Microbiology and Infectious Diseases Department, Hospital General Universitario ‘‘Gregorio Maran˜o´n,’’ Universidad Complutense, Madrid, Spain

INTRODUCTION Solid organ transplant (SOT) recipients may require intensive care unit (ICU) admissions for different reasons in different moments of their evolution, and infection is the most important one. Between 5% and 50% of transplantation candidates must await transplantation in an ICU and, after the procedure, most of them spend a mean of four to seven days there for life support (1–6). If the ICU stay is prolonged due to postsurgical complications, the probability of acquiring a nosocomial infection increases significantly. Most ICU days will take place during the period of deepest immunosuppression (7), but transplant recipients may require readmission to the ICU at any time due to infectious and noninfectious complications such as severe rejection, bleeding, organ dysfunction, etc. In fact, infections are the most common indication for admission of transplant recipients in emergency departments (35%), and severe sepsis (11.7%) is the most common reason for ICU utilization (8). Figures regarding infection and ICU admission show that one-half of all febrile days in liver recipients occur in the ICU, and 87% of these are caused by infection (9). In a multicentric study in Italy, it was shown that most centers are not supported by an ICU exclusively dedicated to transplantation (10). Accordingly, many of these patients will be cared by physicians not always familiar with the specific problems posed by the transplant population. Our aim is to provide information and guidelines regarding most frequently encountered clinical scenarios relevant to critically ill infected SOT recipients. This chapter deals with the etiology, approach, and outcome of most common infectious complications intensive care specialists may 459

Mun˜oz et al.

460

find when taking care of SOT recipients. Where no solid data were available, perspectives based on our own experience and opinion are presented. INFLUENCE OF THE TYPE OF TRANSPLANTATION AND OF THE TIME AFTER TRANSPLANTATION The incidence of infection after a heart transplantation (HT) ranges from 30% to 60% (with a related mortality of 4–15%), and the rate of infectious episodes per patient is 1.73 in a recent series (11). Infections are more frequent and severe than those occurring in renal transplant recipients, but less frequent than those occurring after liver or lung transplantation. The type of SOT and the time after transplantation may be useful clues to the clinician because, unless unexpected exposure has occurred, there is a timetable according to which different infections occur postorgan transplantation (12,13). According to it, although, for example, pneumonia can occur at any point in the posttransplant course, the etiology will be very different at very different points in time. Importance of the Underlying Disease and Type of Transplantation The type of organ transplanted, the degree of immunosuppression, the need for additional antirejection therapy, and the occurrence of technical or surgical complications all impact on the incidence of infection posttransplant. Within each type of transplantation there are patients in which the risk of infection is greater. In HT, patients with prior ischemic cardiomyopathy experience more surgical complications, longer postoperative mechanical assistance, and are more susceptible to Pneumocystis jiroveci pneumonia (14,15) (Table 1). Incidence of infection is higher in pediatric thoracic transplantation than in adult patients (16). After orthotopic liver transplantation (OLT), patients with prior fulminant liver disease fared the worst ICU course and cirrhotics the best (17). Thrombocytopenia Table 1 Risk Factors for Infections in Heart Transplant Patients Preoperative period Pulmonary hypertension not responsive to vasodilators Critically ill status and mechanically ventilated patients at time of transplanation Renal insufficiency Cardiac cachexia Prior sternotomy Donor’s CMV positive serology Older age Repeated hospital admissions Lack of pathogen-specific immunity Latent infections in the donor or the recipient Abbreviations: CMV, cytomegalovirus.

Intraoperative period

Postoperative period

Prolonged operative time Complicated surgical procedure Need for large number of blood transfusions Need for ventricular assist devices Presence of pathogens in the transplant allograft

Prolonged stay in intensive care unit Mediastinal complications and need for reintervention Prolonged hospitalization Prolonged antibiotic use Renal insufficiency Induction therapy with with OKT31 Immunosuppressive drugs and treatment of allograft rejection Immunosuppression due to concomitant viral infections Retransplanation

Infection in Organ Transplant Patients in the Critical Care Unit

461

of < 50  109/L for three days is frequent after liver transplantation and as such was not found to be an important contributor to bleeding. The unique associated event identified for significant bleeding was sepsis (HR, hazard rate 34.80; 95% CI, confidence interval 1.47–153.40) (18). If severely ill patients with end-stage liver disease are selected appropriately, liver transplant outcomes are similar to those observed among subjects who are less ill and are transplanted electively from home (19). Following lung transplantation, patients with obstructive lung disease, double lung transplant, or cystic fibrosis have a longer stay in the ICU and a higher risk of infection (2,20,21). The type of SOT also determines the complexity of the surgery, the intensity of immunosuppression, and the most likely sites of infection. Lung and HT recipients are especially susceptible to thoracic infections, whereas intra-abdominal complications predominate in OLT or pancreas recipients. Patients receiving alentuzumab are more prone to suffer fungal infections (22). Certain infections are characteristic of a particular type of transplantation, e.g., infections related to circulatory support devices (intra-aortic balloon pumps, ventricular assistance devices, and total artificial hearts) in heart transplant recipients (23–25) or endotipsitis in cirrhotic patients (26). Infections such as insertion site sepsis, endocarditis, pneumonia, candidiasis, or sternal infection may complicate 38% of support courses. Lung transplant recipients are admitted to the ICU most commonly due to respiratory deterioration requiring mechanical ventilation (59%) or due to suspicion of sepsis (35%) (27). The use of extended donors does not seem to increase the risk of poor outcome (28). Some characteristics have been found to have a negative impact on liver graft survival (elderly donor with hypertension combined with the presence of metabolic acidosis, or a prolonged ICU donor stay) (29). Time of Appearance of Infection after Transplantation All SOT recipients share a number of conditions (end-stage organ failure, surgery, immunosuppressive regimens, etc.) that bring along a predictable time line of posttransplant infectious complications. The time of appearance of infection after transplantation is an essential component of the evaluation of the etiology of infection. Early infections occurring within the first month after transplantation are generally similar to nontransplant patients who have undergone major surgery in the same body area. Intermediate infections (two to six months) are usually caused by opportunistic microorganisms, such as cytomegalovirus (CMV), fungi, and multiresistant bacteria. Finally, late infections (after six months) may be caused either by common community pathogens in healthy patients or by opportunistic microorganisms in patients with chronic rejection (Table 2). Early Infections In the first month after SOT, patients are very susceptible to ventilator-associated pneumonia, IV catheter-related infections, surgical wound infection, or urinary tract infection (UTI) usually due to bacterial or candidal infections. Some of these may not be evident during the initial examination, which should be frequently repeated. If the patient is still intubated and the chest X ray does not reveal infiltrates, the possibility of tracheobronchitis or bacterial sinusitis should be considered. Staphylococci or enterobacteriaceae will cause most early infections. Gram positives predominate if quinolone prophylaxis is given. Herpetic stomatitis and infections transmitted with the allograft or present in the recipient may also appear at this time.

Mun˜oz et al.

462

Table 2 Chronology of Most Common Infections or Causative Microorganisms in Severely Ill Solid Organ Transplant Recipients Chronology of infection Early infection (first month)

Intermediate infections (2–6 month) Late infections (after sixth month)

Most common syndromes Bacterial infections Pneumonia Surgical wound infection Deep infections near the surgical area Intra-abdominal abscesses Urinary tract infection Catheter-related infection Bloodstream infection Antibiotic associated diarrhea Viral infections Herpes simplex stomatitis HHV-6 infections Primary CMV disease Infections transmitted with the allograft Invasive aspergillosis or candidiasis Opportunistic infections: bacterial, tuberculosis, nocardiosis, invasive aspergillosis, other fungal infections, viral diseases, toxoplasmosis Common community-acquired infections Respiratory tract infections Urinary tract infections Varicella-zoster infections CMV, adenovirus Other opportunistic microorganisms: listeriosis, Cryptococcus, Pneumocystis jiroveci

Abbreviations: HHV-6, human herpesvirus–6; CMV, cytomegalovirus.

Bleeding or anastomosis dehiscences may require a new surgical intervention. Prolonged ICU stay due to central nervous system (CNS) lesions or organ failure usually implies involvement of more resistant species such as vancomycin resistant enterococci (VRE), Acinetobacter, Pseudomonas, methicillin resistant Staphylococcus aureus (MRSA) or Candida (30). Aspergillus may also cause early infection in patients requiring prolonged admission to the ICU and who are especially difficult to diagnose (31). Intermediate Period From the second to the sixth month, patients are susceptible to opportunistic pathogens that take advantage of the immunosuppressive therapy. In this period we may expect infection with immunomodulatory viruses and with opportunistic pathogens (Pneumocystis jiroveci, Listeria monocytogenes, and Aspergillus species). Most lifethreatening infections occur within the first three months. CMV is the most common pathogen after SOT. When no prophylaxis is given, 30% to 90% of patients will show laboratory data of ‘‘CMV infection’’ and 10% to 50% may develop associated clinical manifestations (CMV disease). However, CMV disease is readily diagnosed at present and seldom requires ICU admission. In our experience, only gastrointestinal and respiratory CMV has required ICU admission. Cultures for human herpesvirus (HHV)–6 should be ordered in patients with leukopenia. Some bacterial infections such as listeriosis may appear at this time as primary sepsis or meningitis.

Infection in Organ Transplant Patients in the Critical Care Unit

463

Tuberculosis and nocardiosis are also characteristic of this second period (32). Aspergillosis (IA) may be encountered in patients with risk factors or massive exposure (33) and toxoplasmosis in seronegative recipients of a seropositive allograft (34). Late Period From the sixth month onwards SOT patients are susceptible to community-acquired infections if chronic rejection is not present. Herpes zoster virus, bacterial pneumonia, and UTI predominate. At this time, fever of unknown origin should be managed almost as in immunocompetent hosts. However, the aforementioned opportunistic infections may complicate this late period in patients with chronic viral infection, such as hepatitis B or C, which may progress to end-stage organ dysfunction and/ or cancer. Patients requiring chronic hemodialysis, malignancy, or with late rejection are also susceptible to opportunistic infections (Cryptococcus neoformans, P. jiroveci, L. monocytogenes, etc.) in this timeframe (35). Anamnesis and Physical Examination Risk factors for infection should be carefully sought in all SOT patients admitted to the ICU because they may suggest an etiology and a clinical syndrome. The pretransplantation history, e.g., serological status against microorganisms such as CMV, hepatitis virus, Toxoplasma, etc., may yield valuable information. Previous infections or colonization, exposure to tuberculosis, contact with animals, raw food ingestion, gardening, prior antimicrobial therapy or prophylaxis, vaccines or immunosuppressors, and contact with contaminated environment or persons should be recorded (36,37). History of residence or travel to endemic areas of regional mycosis (38) or Strongyloides stercoralis may be essential to recognize these diseases (39). Exposure to ticks may be essential to diagnose entities such as human monocytic ehrlichiosis, which may be potentially lethal in immunosuppressed patients (40). Diagnosis may be confirmed by polymerase chain reaction (PCR) for Ehrlichia chaffeensis, serology, and by in vitro cultivation of E. chaffeensis from peripheral blood. Certain complications may increase the risk of bacterial and fungal infection in the early posttransplant period. They include long operation (over eight hours), blood transfusion in excess of 3 L, allograft dysfunction, pulmonary or neurological problems, diaphragmatic dysfunction, renal failure, hyperglycemia, poor nutritional state, and thrombocytopenia (17,41–44). Intraoperative hypothermia increased the incidence of early CMV infection in liver transplant recipients (45). Blood cell transfusions have been associated with an increased risk of ventilator-associated pneumonia (46), and leukocyte reduction of all administered blood products during OLT was associated with an improved outcome demonstrated by both a decreased incidence of acute cellular rejection and length of hospital stay (47). Critically ill orthotopic liver transplant patients with kidney failure managed with a conservative anticoagulation policy and continuous venovenous hemofiltration (CVVH) have a much better outcome than acute renal failure (ARF) without orthotopic liver transplantation (OLTX) (48). Fever in critically ill transplant recipients should be considered an emergency. In our opinion, a basic tenet of the management of a SOT with fever is that physical examination data should be directly obtained by the ID consultant, not relying on second hand information. This may be more useful than many expensive and time-consuming tests. The oral cavity is frequently forgotten and may disclose previously unnoticed herpetic gingivo-stomatitis or ulcers. Within the exploration of the thoracic area, the consultant should visualize the entry sites of all intravascular devices, even if they ‘‘have just

464

Mun˜oz et al.

been cleansed.’’ It should be remembered that the presence of inflammatory signs is suggestive of infection, although their absence does not exclude infection. Sepsis, without local signs, may be the initial sign of postsurgical mediastinitis. When the sternal wound remains closed, a positive epicardial pacer wire culture may be a clue to sternal osteomyelitis (49). Although unusual after SOT, cardiac auscultation and echography may help to detect endocarditis (50), and physical examination may occasionally disclose the existence of pneumonia, or empyema before abnormal radiological signs become evident. The abdominal examination is always essential, especially in OLT recipients. The surgical wound is also a common site of infection and a cause of fever. Its presence requires rapid debridement and effective antimicrobial therapy and should prompt the exclusion of adjacent cavities or organ infection. The presence of ascites should be immediately analyzed and properly cultured to exclude peritonitis. We recommend bedside inoculation in blood-culture bottles due to its higher yield of positive results. Examination of the iliac fossa is particularly important after kidney transplantation. Tenderness, erythema, fluctuance, or increase in the allograft size may indicate the presence of a deep infection or rejection. Ultrasound or computed tomography (CT)–guided aspiration may facilitate the diagnosis. The possibility of colonic perforation in steroid-treated patients or gastrointestinal CMV disease should always be considered in intra-abdominal infections. It is important to remember that even very severe intestinal CMV disease may occur in patients with negative antigenemia, especially in patients on mycophenolate mofetil (MMF) (51). Finally, skin and retinal examination are ‘‘windows’’ at which the physician may look in and obtain quite useful information on the possible etiology of a previously unexplained febrile episode. We have analyzed the value of ocular lesions in the diagnosis and prognosis of patients with tuberculosis, bacteremia, and sepsis (52,53). Cutaneous or subcutaneous lesions are a valuable source of information and frequently allow a rapid diagnosis. Viral and fungal infections are the leading causes of skin lesions in this setting. The entire skin surface should be inspected and palpated in SOT recipients with unexplained fever. The biopsy of nodules, subcutaneous lesions, or collections may lead to the immediate diagnosis of invasive mycoses and infections caused by Nocardia or Mycobacteria, among others. An aggressive diagnostic approach is necessary when dealing with febrile compromised ICU hosts because it has been shown or documented that many infectious complications remain undiagnosed. In a recent study, complete agreement between pre- and postmortem diagnoses took place in only 58% of a total 149 patients. Two-thirds of all missed diagnoses were infectious, and disagreement was particularly prominent in the transplant population (complete agreement 17% and major error in 61%) in comparison with trauma patients (complete agreement 86%) or cardiac surgery group (69%). The majority of the missed diagnoses were fungal infections. Longer ICU stays increased the rate of error (31). Approximately 25% of febrile episodes do not present with an evident focal origin and do not permit a straight syndromic approach (54). Therefore, it is essential to know the patient’s antecedents, type of transplantation, and time after surgery. We systematically recommend to our residents to go over the viral, bacterial, fungal, and parasitic etiologies that should be excluded. MOST COMMON CLINICAL SYNDROMES Pneumonia Pneumonia accounts for 30% to 80% of infections suffered by SOT recipients and for a great majority of episodes of fever in the ICU (41% of all febrile infections during

Infection in Organ Transplant Patients in the Critical Care Unit

465

the first seven days of ICU stay and 14% of those after seven days) (9). Pneumonia is among the leading causes of infectious mortality in this population. The incidence of pneumonia is higher in the early postoperative period, especially in the patients who require prolonged ventilation. The clinical presentation and the differential diagnosis are similar to those in other critical patients. The incidence of bacterial pneumonia is highest in recipients of heart–lung (22%) and liver transplants (17%), intermediate in recipients of heart transplants (5%), and lowest in renal transplant patients (1–2%). The crude mortality of bacterial pneumonia in solid organ transplantation has exceeded 40% in most series (55). Pneumonias occur in 13% to 34% of liver transplant recipients. Singh et al. have recently analyzed 40 OLT who developed lung infiltrates in the ICU (35). The etiology was pulmonary edema 40%, pneumonia 38%, atelectasis 10%, acute respiratory distress syndrome (ARDS) 8%, contusion 3%, and unknown 3%. The signs that suggest an infectious origin were clinical pulmonary infection score (CPIS) score > 6 (73% vs. 6%), abnormal temperature (73% vs. 28%), and creatinine level > 1.5 mg/dL (80% vs. 50%) (35). Methicillin resistant Staphylococcus aureus, Pseudomonas aeruginosa, and Aspergillus caused 70% of all pneumonias in the ICU (9). All Aspergillus and 75% of MRSA pneumonias, but only 14% of the gram-negative pneumonias, occurred within 30 days of transplantation. Legionella, Toxoplasma gondii, and CMV may also cause pneumonia in this setting (7,56). Pneumonia is the most common infection following HT. Gram-negative pneumonia in the early posttransplant period is associated with significant mortality. In a recent multicentric prospective study performed in Spain, the incidence of pneumonia after HT was 15.6 episodes/100 HT (57). Most cases occurred in the first month after transplantation. Etiology could be established in 61% of the cases. Bacteria caused 91% of the cases, fungi 9%, and virus 6%. In another study, opportunistic microorganisms caused 60% of the pneumonias, nosocomial pathogens 25%, and communityacquired bacteria and mycobacteria 15% (58). Gram-negative rods caused early pneumonias (median nine days), gram-positive cocci (11 days), fungi (80 days), Mycobacterium tuberculosis and Nocardia spp. (145 days), and virus (230 days). Legionella should always be included in the differential diagnosis (59–62). Pneumonia increases the risk of mortality after HT (odds ratio (OR) 3.7, IC 95% 1.5–8.1, P < 0.01). Lung infections are very common in lung and heart–lung transplant recipients. These patients have particular predisposing factors because the allograft is in contact with the outside environment, and have an impaired mucociliary clearance, ischemic lymphatic interruption, and abolition of the cough reflex distal to the tracheal or bronchial anastomoses. In fact, the anastomosis is especially vulnerable to invasion with opportunistic pathogens including gram-negative bacilli (Pseudomonas), staphylococci, or fungus. Lung transplant recipients with underlying cystic fibrosis may be prone to suffer infections caused by multiresistant microorganisms such as Burkholderia cepacia. In this group of patients perioperative antimicrobials are chosen on the basis of surveillance cultures. Pathogens transmitted from the donor may also cause pneumonia in this setting. Pneumonia is less common after renal transplantation (8–16%), although it remains a significant cause of morbidity (63–65). Most Common Pathogens in Transplant Patients with Pneumonia We have already mentioned some data on the etiology of pneumonia in SOT recipients, but we will now review in more detail some of the most common groups of pathogens. Bacteria. Although bacterial pneumonia may occur any time after transplantation, the period of greater risk is the first month after the procedure. Need for

Mun˜oz et al.

466

Table 3 Probable Etiology of Pneumonia in Relation to the Type and Progression of the Infiltrates Radiologic pattern Consolidation

Interstitial Nodular

Acutea

Subacute

Bacteria (Streptococcus pneumoniae gram-negative rods, Legionella, staphylococci) Embolisms Hemorrhage CMV Edema, Transfusions (Bacteria) (Bacteria, edema)

Aspergillus, Nocardia, tuberculosis, drugs, Pneumocystis jiroveci, Legionella, HSV, VVZ, Toxoplasma

Virus, P. jiroveci, drugs (Fungi, Nocardic, tuberculosis) Fungi, Nacardia, tuberculosis (P. jiroveci, CMV )

a Acute: require attention in < 24 hr. Less common possibilities are among brackets. Abbreviations: CMV, cytomegalovirus; HSV, herpes simplex virus; VVZ, virus waucella zoster.

mechanical ventilation and intensive care in this period are among the causes. The etiology will depend on the moment after transplantation, length of previous hospital stay, the days on ventilation, previous use of antimicrobial agents, and clinical and radiological manifestations (Table 3). Gram-negative rods predominate (P. aeruginosa, Acinetobacter spp., and Enterobacteriaceae) but gram-positive cocci (S. aureus, S. pneumoniae) account for a significant proportion of cases, as we mentioned before. Legionella has been reported in 2% to 27% of SOT recipients with pneumonia (66–68). Most common species implicated are Legionella pneumophila and L. micdadei (69,70). A prodrome of influenza-like symptoms is followed by a sometimes ‘‘explosive’’ pneumonia with patchy lobular or interstitial infiltrates on chest radiograph. High fever, hypothermia, abdominal pain, and mental status changes are sometimes seen. Pneumonia is the most common presentation, but some patients have just fever (62). Other manifestations have also been described such as liver abscesses, pericarditis, cellulitis, peritonitis, or hemodialysis fistula infections (71). Infiltrate is usually lobar, but Legionella has to be included in the differential diagnosis of lung nodules, cavitating pneumonia, and lung abscess (59). Legionella infections can be overlooked unless specialized laboratory methodology (cultures on selective media, urinary antigen) is applied routinely on all cases of pneumonia (60). Routine culture of the water supply for Legionella is recommended in all transplant centers and ICUs with cases of Legionellosis (72). The use of impregnated filter systems may help prevent nosocomial Legionellosis in high-risk patient care areas (73). The frequency of M. tuberculosis disease in receptors of solid organ transplantation in most developed countries ranges from 1.2% to 6.4%, but in transplant patients living in areas of high-level endemicity it might reach up to 15% (32,74–76). Although there is a huge regional variability, in general SOT incidence is 20 to 74 times higher than in the general population, with a mortality rate of up to 30%. The most frequent form of acquisition of tuberculosis after transplantation is the reactivation of latent tuberculosis in patients with previous exposure. Tuberculosis develops a mean of nine months after transplantation (0.5–13 months). Risk factors for early onset are nonrenal transplant, allograft rejection, immunosuppressive therapy with

Infection in Organ Transplant Patients in the Critical Care Unit

467

(OKT31) anti-CD3 monoclonal antibodies or anti-T cell antibodies, and previous exposure to M. tuberculosis. Clinical presentation is frequently atypical and diverse, with unsuspected and elusive sites of involvement. A large series of tuberculosis (TBC) in transplant recipients described pulmonary involvement in 51% of patients, extrapulmonary tuberculosis in 16%, and disseminated infection in 33% (32). In lungs, radiographic appearance may vary between focal or diffuse interstitial infiltrates, nodules, pleural effusion, or cavitary lesions. Manifestations include fever of unknown origin, allograft dysfunction, gastrointestinal bleeding, peritonitis, or ulcers. In transplant patients, M. tuberculosis infection was also described in skin, muscle, osteoarticular system, CNS, genitourinary tract, lymph nodes, larynx, adrenal glands, and thyroid (32,77). Ocular lesions may be an early way to detect dissemination (52). Coinfection with other pathogens is not uncommon. Treatment requires control of interactions between antituberculous drugs and immunosuppressive therapy. A high index of suspicion is recommended. Rhodococcus equi (78) and Nocardia (79–83) are well-known causes of respiratory tract infection in transplant recipients. However, they usually present in a subacute form and rarely require ICU admission. These infections usually occur more than three months posttransplantation. Radiologically, they may appear as multiple and bilateral nodules, possibly due to their long-term silent evolution. The incidence of nocardiosis has been significantly reduced since the widespread use of cotrimoxazole prophylaxis. Nocardia farcinica may be resistant to cotrimoxazole prophylaxis and cause particularly aggressive disease (79). R. equi is an opportunistic pathogen, which usually causes cavitated pneumonia in HIV-positive patients, but SOT recipients may be affected as well. Infection occurs usually late (median of 49 months after transplantation), and the lungs are primarily involved in most cases. Infection presents as a lung nodule in half of the patients. Clinicians should consider R. equi when evaluating a solid organ recipient with an asymptomatic lung nodule, particularly when cultures fail to identify Mycobacteria, Nocardia, or fungal organisms. Clinical microbiology laboratories should be alerted when a R. equi infection is suspected, because it could be mistaken for a contaminant diphtheroid and will not respond to the standard empirical therapy. Fungal infections have been reported to occur in 5% to 20% of SOT recipients, and although they are decreasing proportionally, they increase in absolute figures as more transplantation procedures are performed each year. Rates vary according to the type of transplant recipient and are greatly influenced by the degree of immunosuppression, the use of prophylaxis, the rate of surgical complications, and rate of renal failure among the transplant population. Fungal pathogens more likely to cause pneumonia in this population are Aspergillus, P. jiroveci, Candida spp., and Cryptococcus spp. Different types of transplantations imply differences in fungal infections (84). A recent series prospectively collected in Spain reported the incidence of invasive IA in SOT recipients, which ranged from 0.3% in kidney transplant to 3.9% in pancreas recipients (85). In lung and heart–lung transplantation, the incidence of fungal infections, most notably IA, ranged from 14% to 35% if no prophylaxis was provided, but has significantly decreased because aerosolized amphotericin B is given to these patients (86,87). In single lung transplant patients, invasive IA more commonly affects the native lung than the transplanted lung and may arise immediately postoperatively due to preexistent disease in pretransplant immunosuppressed patients. In lung and heart–lung transplant recipients the types of disease presentation include bronchial anastomosis dehiscence, vascular anastomosis erosion, bronchitis, tracheobronchitis, invasive lung disease, aspergilloma, empyema,

468

Mun˜oz et al.

disseminated disease, endobronchial stent obstruction, and mucoid bronchial impaction. Kramer et al. have described a distinct form of IA after lung transplantation: ulcerative tracheobronchitis, a semi-invasive disease involving the anastomosis site, and the large airways (88). Risk factors include CMV infection, obliterative bronchitis, rejection, and increased immunosuppression. In HT, Aspergillus is the predominant fungal isolate and accounts for 38% of all lung nodular lesions (89). It appears a median of 50  63 days after HT (90). We found that postoperative hemodialysis, CMV disease, reoperation, and other episodes of IA in the ward close to the transplantation date are the major risk factors for IA in this population. The use of oral itraconazole is an effective way of preventing this infection. In liver transplantation, Aspergillus infection is less common when compared to lung or heart–lung transplant recipients, but is more commonly found than in kidney transplant recipients. In liver transplant recipients, IA usually is an early event, and most patients were still in the ICU with evidence of organ dysfunction when the disease was diagnosed (76,91). Retransplantation is also an independent risk factor (91,92), although IA may happen in low-risk patients if an overload exposure has occurred (33). Accordingly, ICUs caring for transplant patients should maintain a good quality of air control (93). Aspergillus may appear late after transplantation, mainly in patients with a neoplastic disease (94). Pulmonary involvement is described in 90% of the cases, but CNS or disseminated manifestations may predominate (95). The isolation of Aspergillus from any SOT recipient sample is always a warning clue. Although the lung is the primary site of infection, other presentations have also been described (surgical wound, primary cutaneous infection, infection of a biloma, endocarditis, endophthalmitis, etc.). Scedosporium species are increasingly recognized as significant pathogens, particularly in immunocompromised hosts. These fungi now account for 25% of all non-Aspergillus mold infections in organ transplant recipients (96). Scedosporium species are generally resistant to amphotericin B. Scedosporium prolificans, in particular, is also resistant to most currently available antifungal agents. We found that 46% of Scedosporium infections in organ transplant recipients were disseminated and that patients may occasionally present with shock and sepsis-like syndrome (97). Fungemia is especially frequent when S. prolificans is involved. Overall, mortality rate for Scedosporium infections in transplant recipients in our study was 58%. When adjusted for disseminated infection, voriconazole as compared to amphotericin B was associated with a lower mortality rate that approached statistical significance (p ¼ 0.06). P. jiroveci (former P. carinii) is now rarely seen in SOT receiving prophylaxis. Before prophylaxis, incidence was around 5%, although it has been described to reach up to 80% in lung transplant recipients. P. jiroveci pneumonia was diagnosed a median of 75 days after transplant (range, 37–781 days). Clinical presentation was acute (less than 48 hours) with fever (89%), shortness of breath (84%), dry cough (74%), and hypoxia (63%). CMV was isolated from lung or blood in 74% of patients. Chest X ray usually showed interstitial pneumonia (84%). Some patients required ventilatory support. Mortality was 26%. Older age was the only significant poor prognostic factor (61 years vs. 49 years; p < 0.03) (15). Weekend prophylaxis (one double-strength tablet, 160/800 mg, every 12 hours on Saturdays and Sundays) has shown practically universal efficacy, also eliminating cases of Listeria or Nocardia infections. C. neoformans affects the lung in 55% of SOTs with cryptococcosis (98). However, the disease is uncommon and appears a median of 24 months after

Infection in Organ Transplant Patients in the Critical Care Unit

469

transplantation (1 month to 17 years). An immune reconstitution syndrome–like entity may occur in organ transplant recipients with C. neoformans infection. This entity may be interpreted as failure of therapy. Immunomodulatory agents may have a role as adjunctive therapy in such cases (99). Although Candida is frequently recovered from the lower respiratory tract of ventilated patients, Candida pneumonia is exceedingly rare (100). It has been reported in lung transplant recipients, and the diagnosis requires histological confirmation, because the recovery of Candida may represent colonization. In these patients, infection with Candida may be associated with very severe complications such as the necrosis of bronchial anastomoses (101–104). CMV was the most common organism infecting the lungs in solid transplant recipients, but the incidence has significantly decreased with the widespread use of prophylaxis. CMV may be the sole causative agent of pneumonia after SOT or appear as a copathogen when other microorganisms are isolated (61). CMV pneumonitis commonly adopts a diffuse interstitial radiological appearance, but focal and even nodular infiltrates are described in up to one-third of patients. CMV may cause severe pneumonia with ARDS requiring ICU admission. In a recent study, in kidney transplant recipients, including 21 patients in this situation, it was found that among 13 surviving patients, the numbers of CD4þ and CD8þ T cells and their ratio increased as the patients recovered. In eight nonsurviving patients, the numbers of CD4þ and CD8þ T cells and their ratio was similar to day 0. It was concluded that the variations of CD4þ and CD8þ T lymphocytes and their ratio are useful indicators of the severity of disease and the outcome of patients with CMV infections accompanying ARDS after renal transplantation. Nevertheless, it may be helpful to evaluate the efficiency of ongoing treatment methods in these patients.(105) Herpes simplex (106,107) and virus vamcella zoster (VVZ) may also cause pneumonia in the transplant population. human herpes virus 6 (HHV)-6 has been reported to cause diverse clinical symptoms including fever, skin rash, pneumonia, bone marrow suppression, encephalitis, and rejection. The respiratory viruses, particularly respiratory syncytial virus, influenza, parainfluenza, adenovirus, and picornaviruses, are increasingly recognized as significant pathogens in these populations. Adenovirus may also cause pneumonia, occasionally with dysfunction of the allograft (108). Respiratory syncytial virus and influenza have been found to be the most common of the respiratory viruses causing severe infections in transplant recipients (109–115). New antiviral medications may bring improved outcomes of picornavirus infections in this population. Finally, a new virus, the human metapneumovirus, has recently been described and may be a significant respiratory pathogen in immunocompromised transplant recipients (116). Respiratory viruses may be associated with high morbidity, particularly in lung transplant recipients, and may appear as ‘‘culture-negative’’ pneumonia. Molecular methods such as reverse transcription-PCR assays allow the identification of respiratory viruses in bronchoalveolar lavage (BAL) specimens (117). Advances in prevention, particularly with regard to infection control practices, and to a lesser extent treatment, have had a substantial impact on the frequency and outcomes of this infection. Considering the high mortality that some of these pathogens engender, the prompt detection of the etiology is of the utmost importance. As with other critical patients, differentiating pneumonia from other etiologies of pulmonary infiltrates can be extremely difficult. In liver transplant patients, a CPIS score > 6, abnormal temperature, and renal failure (serum creatinine > 1.5 mg/dL) were significant predictors of pneumonia (35). It is important to bear in mind that some drugs, such as

470

Mun˜oz et al.

sirolimus, may cause pulmonary infiltrates (118). Patients may develop dyspnea, cough, fatigue, and sometimes fever. Characteristic radiological changes are bilateral lower zone haziness. The presentation ranges from insidious to fulminant, and usually there is a rapid response to sirolimus withdrawal. Chest X rays of transplant recipients with pneumonia predominantly show alveolar or interstitial infiltrates of variable extension. However, nodular lesions are not uncommon. The differential diagnosis of a lung nodule in a normal host includes many malignant and benign processes. However, in immunosuppressed patients the most common causes are potentially life-threatening opportunistic infections that may be treated and prevented. We have detected single or multiple lung nodules on the chest radiograph in 10% of our HT patients (89). Aspergillus infection was detected early after transplantation (median 38 days, range 23–158), whereas Nocardia asteroides and Rhodococcus infections developed only later (median 100 days, range 89–100). Nodules due to CMV occurred 16 to 89 days after HT (median 27 days). Patients with Aspergillus were, overall, more symptomatic and were the only ones in our series to present neurological manifestations and hemoptysis. CT is more sensitive than standard chest X ray in identifying the number of lesions and may assist guided biopsy. Etiologic diagnosis is mandatory considering that only 50% of the empirical treatments of pneumonia in HT patients are appropriate (58). For this reason, fast diagnostic procedures that guide antimicrobial treatment are necessary. Etiologic diagnosis may be performed by using different techniques, so this requires careful tailoring to each single patient. Once pneumonia is identified, blood cultures, respiratory samples for culture of bacteria, mycobacteria, fungi and viruses, and urine for Legionella and S. pneumoniae antigen detection must be sent to the laboratory (if possible, before starting antimicrobials). The rate of expected bacteremia in patients with pneumonia is 16% to 29% (119). Demonstration of pathogenic microorganisms (M. tuberculosis, Legionella, Cryptococcus spp., R. equi, or P. jiroveci) in a sputum sample is diagnostic. PCR techniques may help improve diagnostic sensitivity (74). A bronchoscopic sample with bronchial biopsy is preferable for CMV, Aspergillus, or P. jiroveci pneumonia. If pleural fluid is present it should also be analyzed. In our series of nodular lesions in HT patients, etiological diagnosis was established within a median of eight days (1–24). A median of 1.8 invasive techniques per patient was necessary to achieve the diagnosis. Overall diagnostic yield was 60% for transtracheal aspiration, 70% for BAL, and 75% for transthoracic aspiration. BAL was the first positive technique in 58% of the patients. The only complications were a minor pneumothorax after a transbronchial biopsy and minor hemoptysis after a transthoracic needle aspiration. Direct microscopic examination of the respiratory samples (Gram stain, potassium hydroxide, or cotton blue preparations) was positive in three out of five cases of IA and in three out of four cases of Nocardiosis (89). A serum sample should also be submitted. Pneumonia is the infection with the highest related mortality rate, and this is also true for SOT recipients, so prompt empirical therapy is highly recommended for patients in critical condition after obtaining adequate samples. The selection of the empirical therapy will be guided by the characteristics of the patient and the clinical situation. Postsurgical Infections Complications in the proximity of the surgical area must always be investigated. Surgical problems leading to devitalized tissue, anastomotic disruption, or fluid

Infection in Organ Transplant Patients in the Critical Care Unit

471

collections markedly predispose the patient to potentially lethal infection. In the early posttransplantation period, renal and pancreas transplant recipients may develop perigraft hematomas, lymphoceles, and urinary fistula. Liver transplant recipients are at risk for portal vein thrombosis, hepatic vein occlusion, hepatic artery thrombosis, and biliary stricture formation and leaks. Heart transplant recipients are at risk for mediastinitis and infection at the aortic suture line, with resultant mycotic aneurysm, and lung transplantation recipients are at risk for disruption of the bronchial anastomosis. Intra-abdominal Infection In OLT recipients intra-abdominal infections may be responsible for 50% of bacterial complications and cause significant morbidity (120); they include intra-abdominal abscesses, biliary tree infections, and peritonitis. In nonabdominal transplantations, intra-abdominal infections may be caused by preexisting problems such as biliary tract lithiasis, diverticulitis, CMV disease, etc. Risk factors for intra-abdominal complications after OLT include prolonged duration of surgery, transfusion of large volumes of blood products, use of a choledochojejunostomy (rous-en-Y) instead of a choledochostomy (duct-to-duct) for biliary anastomosis, repeat abdominal surgery of the biliary tract, dehiscence or obstruction, intra-abdominal hematomas, vascular problems of the allograft (for example the thrombosis of the hepatic artery or the ischemia of the biliary tract may create the apparition of cholangitis and liver abscesses), and CMV infection. Occasionally, the complications will appear after the performance of some procedure such as a liver biopsy or a cholangiography. These infections may be bacteremic, and in fact, OLT recipients show the highest rate of secondary bloodstream infections. Most common microorganisms include Enterobacteriaceae, enterococci, anaerobes, and Candida. In a series published by Singh et al. biliary tree was the origin of 9% of infections associated with fever in the ICU (9). Biliary anastomosis leaks may result in peritonitis or perihepatic collections, cholangitis, or liver abscesses (121–123). OLT recipients are especially predisposed to suffer cholangitis. Recent data suggest that duct-to-duct biliary anastomosis stented with a T-tube tends to be associated with more postoperative complications (124). A percutaneous aspirate with culture of the fluid is required to confirm infection. Culture of T-tube is unreliable because it may only reflect colonization. Hepatic abscess is frequently associated with hepatic artery thrombosis (125). In one series, median time from transplant to hepatic abscess was 386 days (range 25–4198). Clinical presentation of hepatic abscess was similar to that described in nonimmunosuppressed patients. Occasionally the only manifestations are unexplained fever and relapsing subacute bacteremia. In fact 40% to 45% of the liver abscesses are associated with bacteremia. Prolonged antibiotic therapy, drainage, and even retransplantation may be required to improve the outcome in these patients. Catheter drainage was successful in 70% of cases. Mortality rate was 42% (126). Ultrasonography and CT of the abdomen are the normal techniques to identify intra-abdominal or biliary infections. However, sterile fluid collections are exceedingly common after liver transplantation, so an aspirate is necessary to establish infection. Mediastinitis In heart and lung transplant recipients the possibility of mediastinitis (2–9%) should be considered. HT patients have a higher risk of postsurgical mediastinitis and

472

Mun˜oz et al.

sternal osteomyelitis than other heart surgical patients (127). It may initially appear merely as fever or bacteremia of unknown origin. Inflammatory signs in the sternal wound, sternal dehiscence, and purulent drainage may appear later. The most commonly involved microorganisms are staphylococci, but gram-negative rods represent at least a third of the cases. Mycoplasma, mycobacteria, and other less common pathogens should be suspected in ‘‘culture-negative’’ wound infections (Thaler, 1992 #7537; Levin, 2004 #5135). A bacteremia of unknown origin during the first month after HT should always suggest the possibility of mediastinitis. Risk factors are prolonged hospitalization before surgery, early chest reexploration, low output syndrome in adults, and the immature state of immune response in infants. Therapy consists of surgical debridement and repair, and antimicrobial therapy given for three to six weeks. Urinary Tract Infections Urinary tract infections are the most common form of bacterial complication affecting renal transplant recipients (128,129). The incidence in patients not receiving prophylaxis has been reported to vary from 5% to 36% in recent series (130,131). However, it is not a common cause of ICU admission. The most common pathogens include Enterobacteriaceae, enterococci, staphylococci, and Pseudomonas. However, other less frequent microorganisms, like Salmonella, Candida, or Corynebacterium urealyticum pose specific management problems in this population. It is also important to remember the possibility of infection caused by unusual pathogens such as Mycoplasma hominis, M. tuberculosis, or BK and JC viruses. Unless another source of fever is readily apparent, any febrile kidney transplant patient with an abrupt deterioration of renal function should be treated with empiric antibacterial therapy aimed at gram-negative bacteria, including P. aeruginosa, after first obtaining blood and urine cultures (132). Prolonged administration of antimicrobial therapy has been classically recommended for the treatment of early infections, although no doubleblind, comparative study is available (128). Gastrointestinal Infections Gastrointestinal symptoms are present in up to 51% of HT patients in recent series, although only 15% are significant enough to warrant endoscopic, radiologic, or surgical procedures. Possible manifestations include gastrointestinal bleeding, diarrhea, abdominal pain, jaundice, nausea or vomiting, odynophagia, or dysphagia. Hepatobiliary, peptic ulcer, and pancreatic complications are the most prevalent. Peritonitis, intra-abdominal infections, and Clostridium difficile colitis accounted for 5% of all febrile episodes in OLT in the ICU (9). Abdominal pain and/or diarrhea are detected in up to 20% of organ transplant recipients (119). CMV and C. difficile are the most common causes of infectious diarrhea in SOT patients. CMV may involve the whole gastrointestinal tract, although duodenum and stomach are the most frequent sites involved (133). Infection of the upper gastrointestinal tract with CMV used to be a major cause of morbidity in transplant patients (134). In one series 53 out of 201 heart transplant patients had persistent upper gastrointestinal symptoms (abdominal pain, nausea, and vomiting). Of these 53 patients, 16 (30.2%) had diffuse erythema or ulceration of the gastric mucosa (14), esophagus (1), and duodenum (1) with biopsy results that were positive for CMV on viral cultures (incidence, 8%). All patients with positive biopsy results were treated

Infection in Organ Transplant Patients in the Critical Care Unit

473

with intravenous ganciclovir. Recurrence developed in six patients (37.5%) and required repeated therapy with ganciclovir. None of the 16 patients died as a result of gastrointestinal CMV infection. Other possible presentation symptoms are fever and gastrointestinal bleeding. Differential diagnosis should include diverticulitis, intestinal ischemia, cancer, and Epstein-Barr Virus (EBV)-associated lymphoproliferative disorders. A particular gastric lymphoma called mucosa-associated lymphoid tissue lymphoma may develop in renal transplant patients. It usually responds to the eradication of Helicobacter pylori (135). PCR is an accurate method for the detection of CMV in the mucosa of the gastrointestinal (GI) tract (136). The natural history of CMV disease associated with solid organ transplantation has been modified as a result of the widespread use of potent immunosuppressants and antiviral prophylaxis, and late severe forms are now detected (137). Hypogammaglobulinemia may also justify severe or relapsing forms of CMV after solid organ transplantation (138). C. difficile should be suspected in patients who present with nosocomial diarrhea. It is more common in transplant populations who frequently receive antimicrobial agents, and up to 20% to 25% of patients may experience a relapse (139–141). Incidence of C. difficile infection is increasing, even taking into account improved diagnosis and increased awareness. Most infections occur early after transplantation (140). The most important factor in the pathogenesis of disease is exposure to antibiotics that disturb the homeostasis of the colonic flora. Nosocomial transmission has also been described. SOT recipients have many risk factors for developing C. difficile-associated diarrhea (CDAD): surgery, frequent hospital admissions, antimicrobials exposure, and immunosuppression. Most common clinical presentation is diarrhea, but clinical presentation may be unusually severe (142,143). In a recent series 5.7% of the kidney or pancreas transplant recipients developed fulminant CDAD that presented with toxic megacolon, and underwent colectomy. One of them died; the other patient survived after colectomy (144). Absence of diarrhea is a poor prognostic factor. In these cases significant leukocytosis may be a very useful clue. The infection may be demonstrated with a rectal swab. Occasionally patients present with an acute abdomen (145) or inflammatory pseudotumor (146). The reference method for diagnosis is the cell culture cytotoxin test, which detects the presence of toxin B in a cellular culture of human fibroblasts (147), but recovering C. difficile in culture allows the performance of a ‘‘second-look’’ cell culture assay that enhances the potential for diagnosis (148). CDAD may pose important diagnostic problems in the transplant setting. Clinical presentation may be atypical and sometimes quite severe, differential diagnosis with other entities causing diarrhea in this population is required (CMV, adenovirus), and relapses may be difficult to manage. C. difficile colitis may occur in coincidence with CMV gastrointestinal infection, which may complicate the diagnosis (139). The first step in managing diarrhea and colitis caused by C. difficile is discontinuation of the antibiotic therapy that precipitated the disease, whenever possible. About 15% to 25% of patients respond within a few days. Patients with severe disease should be treated with oral metronidazole or vancomycin. Oral metronidazole (500 mg tid or 250 mg every six hours) and oral vancomycin (125 mg every six hours) administered for 10 to 14 days have similar therapeutic efficacy, with response rates near 90% to 97%. When oral administration is not feasible, IV metronidazole should be used, because IV vancomycin is not effective. Nearly all patients respond to treatment in about five days. Comparison of metronidazole’s activity with that of

474

Mun˜oz et al.

vancomycin in patients with moderately severe disease shows similar response rates. The former is preferred because of its reduced risk of vancomycin-resistance induction and lower cost. However, recent reports of very severe clinical forms suggest that vancomycin may be preferable for these especially virulent strains. C. difficile strains resistant to metronidazole and with intermediate resistance to vancomycin have been described. The administration of probiotics such as Saccharomyces boulardii, or Lactobacillus sp. for prophylaxis of CDAD remains controversial, and we do not recommend it in critical patients because the occurrence of severe invasive disease by S. boulardii has been described (149). As mentioned, a substantial proportion of patients (10–25%) have a relapse usually 3 to 10 days after treatment has been discontinued, even with no further antibiotic therapy. Relapse usually results from either a failure to eradicate C. difficile spores from the colon or due to reinfection from the environment. Nearly all patients respond to another course of antibiotics if given early. The frequency of relapses does not seem to be affected by the antibiotic selected for treatment, the dose of these drugs, or the duration of treatment. Multiple relapses may be difficult to manage. Several measures have been suggested: gradual tapering of the dosage of vancomycin over one to three months, administration of ‘‘pulse-dose’’ vancomycin, use of anion-exchange resins to absorb C. difficile toxin A, administration of vancomycin plus rifampin, or administration of immunoglobulins. Infectious enteritis is especially frequent in intestinal transplant recipients (39%). Viral agents are the cause in two-thirds of the cases. In a recent series there were 14 viral enteritis (one CMV, eight rotavirus, four adenovirus, and one Epstein–Barr virus), three bacterial (C. difficile), and three protozoal infections (one Giardia lamblia and two Cryptosporidium). The bacterial infections tended to present earlier than the viral infections, and the most frequent presenting symptom was diarrhea (150). Immunosuppressive drugs, such as MMF, cyclosporine A, tacrolimus, and sirolimus, are all known to be associated with diarrhea. Rarely, graft-versus-host disease, lymphoproliferative disorder, de novo inflammatory bowel disease (IBD), or colon cancer may present as diarrhea. Flare-up of preexisting IBD is also not uncommon after LT. However, the cause of acute diarrhea remains unidentified in one of three patients (151). Neurological Focality The detection of CNS symptoms in a SOT recipient should immediately arouse the suspicion of an infection (152). Fever, headache, altered mental status, seizures, focal neurological deficit, or a combination of them should prompt a neuroimaging study (119). Noninfectious causes include immunosuppressive-associated leukoencephalopathy (153), toxic and metabolic etiologies, and stroke and malignancies (154). Therapy with OKT3 monoclonal antibody has been related to the production of acute aseptic meningitis [cerebrospinal fluid (CSF) pleocytosis with negative cultures, fever, and transient cognitive dysfunction]. Infectious progressive dementia has been related to JC virus, Herpes simplex, CMV, and EBV. The most common cause of meningoencephalitis in organ transplant recipients is herpes viruses, followed by L. monocytogenes, C. neoformans, and T. gondii. HHV6 is a neurotropic ubiquitous virus known to cause febrile syndromes and exanthema subitum in children. Less commonly, and particularly in organ transplant recipients,

Infection in Organ Transplant Patients in the Critical Care Unit

475

it may cause hepatitis, bone marrow suppression, interstitial pneumonitis, and meningoencephalitis (155). In a recent review, HHV-6 encephalitis occurs a median of 45 days (range 10 days to 15 months) after transplantation. Mental status changes, ranging from confusion to coma (92%), seizures (25%), and headache (25%) were the predominant clinical presentations. Focal neurologic findings were present in only 17% of the patients. Twenty-five percent of the patients had fever, occasionally reaching 40 C. CSF pleocytosis was generally lacking. Magnetic resonance images of the brain may reveal multiple bilateral foci of signal abnormality (nonenhancing involving both gray and white matter). HHV-6 can be detected in CSF by PCR or by viral isolation. HHV-6 viremia was documented in 78% of the patients. Overall mortality in patients with HHV-6 encephalitis was 58% (7 of 12); 42% (5 of 12) of the deaths were caused by HHV-6. Cure was documented in seven of eight patients who received ganciclovir or foscarnet for seven days, compared with 0% (zero of four) in those who did not receive these drugs or received them for < seven days (P ¼ 0.01) (156). A growing body of evidence suggests that the more important effect of HHV-6 and HHV-7 reactivation on the outcomes of liver transplantation may be mediated indirectly by their interactions with CMV (157). HHV-6 viremia is an independent predictor of invasive fungal infection (158). Cytomegalovirus infection of the CNS is quite uncommon in SOT recipients. It may affect the brain (diffuse encephalitis, ventriculoencephalitis, and cerebral mass lesions) or the spinal cord (transverse myelitis and polyradiculomyelitis). Diagnosis is very difficult and should be based on clinical presentation, results of imaging, and virological markers. The most specific diagnostic tool is the detection of CMV DNA by PCR in the CSF. Treatment should be initiated promptly if CMV infection is suspected. Antiviral therapy consists of intravenous ganciclovir, intravenous foscarnet, or a combination of both. Cidofovir is the treatment of second choice. Patients who experience clinical improvement or stabilization during induction therapy should be given maintenance therapy (159). Encephalitis caused by herpes simplex virus (HSV) has also been described (160,161). Among causes of encephalitis, West Nile virus has emerged as an important cause of several outbreaks of febrile illness and encephalitis in North America over the past few years. In a recent report 11 transplant recipients with naturally acquired West Nile Encephalitis (WNE) were identified (four kidney, two stem cell, two liver, one lung, and two kidney/pancreas). Ten patients developed meningoencephalitis, which in three cases was associated with acute flaccid paralysis. All patients had CSF pleocytosis and WNV-specific immunoglobulin M in the CSF and/or serum. Magnetic resonance images of the brain were abnormal in seven of eight tested patients, and electroencephalograms were abnormal in seven of seven, with two showing periodic lateralized epileptiform discharges. Nine of 11 patients survived infection, but three had significant residual deficits. This viral infection should be considered in all transplant recipients who present with a febrile illness associated with neurological symptoms (162–164). L. monocytogenes infections can occur at almost any time, although the most common occurrence is two to six months posttransplant (165). The incidence has significantly been reduced because prophylaxis with cotrimoxazole is used (15). Listeria infections may present as isolated bacteremia or with associated meningitis (166,167). OLT recipients may present with acute hepatitis (168). Brainstem encephalitis or rhomboencephalitis have been characteristically described in patients with Listeriosis, in which cranial nerve palsies or pontomedullary signs may be observed. Cerebritis/abscess due to L. monocytogenes, without meningeal involvement, is less common (169).

476

Mun˜oz et al.

Incidence of cryptococcosis after organ transplantation is 0.3% to 6% (170–172). Cryptococcus is mostly a cause of meningitis, pneumonia, and skin lesions (173–176). However, more uncommon sites of infection have been also described in immunocompromised patients such as hepatic cryptococcosis in a heart transplant recipient (177). The patient developed fever, dyspnea, and signs of liver damage. Diagnosis was made with liver biopsy and with cryptococcal antigen in serum (177). Cryptococcosis is usually a late disease after transplantation, although rare fulminant early cases have been reported (178). CSF analysis usually reveals moderate pleocytosis. CSF cryptococcal antigen is positive in most patients. In a recent series 83 transplant recipients with cryptococcosis were analyzed. Patients with central nervous system infection (69% vs. 16%, P ¼ 0.00001), disseminated infection (82.7% vs. 20%, P ¼ 0.00001), and fungemia (29% vs. 8%, P ¼ 0.046) were more likely to receive regimens containing amphotericin B than fluconazole as primary therapy. Survival at six months tended to be lower in patients whose CSF cultures at two weeks were positive compared to those whose CSF cultures were negative (50% vs. 91%, P ¼ 0.06) (98). Focal brain infection (seizures or focal neurologic abnormalities) may be caused by Listeria, T. gondii, fungi (Aspergillus, Mucorales, phaeohyphomycetes, or dematiaceous fungi), posttransplantation lymphoproliferative disease or Nocardia. Brain abscesses are relatively uncommon (0.6%) in SOT patients, and most of them (78%) are caused by Aspergillus (179), followed by T. gondii and N. asteroides. Aspergillus brain abscesses usually occur in the early posttransplantation period. Most of the patients present with simultaneous lung lesions that allow an easier diagnostic way. Overall, disseminated Aspergillus disease has been described in 9% to 36% of kidney recipients, 15% to 20% of lung recipients, 20% to 35% of heart recipients, and 50% to 60% of liver recipients with IA (95,180). Disseminated infection with CNS involvement occurred in 17% of the cases studied in Spain. Clinical manifestations of CNS IA include alteration of mental status, diffuse CNS depression, seizures, evolving cerebrovascular accidents, and headache (95,181). The CSF is almost always sterile. Toxoplasmosis was more prevalent when prophylaxis with cotrimoxazole was not provided (34,182). The incidence is higher in heart transplant recipients. The disease usually occurred within three months posttransplantation, with fever, neurological disturbances, and pneumonia as the main clinical features. Chorioretinitis may also be found (183,184). Diagnosis was established by serology and by direct examination, culture, or PCR of biological samples. In heart transplant recipients the diagnosis may be provided by the endomyocardial biopsy (185). The lesions of T. gondii are usually multiple, have preferential periventricular localization, and demonstrate ring enhancement. The donor was the likely source of transmission to most recipients (186). The mortality rate was high (around 60%). Obstructive urinary tract lithiasis involving sulfadiazine crystals has been described (187). Disseminated toxoplasmosis should be considered in the differential diagnosis of immunocompromised patients with culture-negative sepsis syndrome, particularly if combined with neurologic, respiratory, or unexplained skin lesion (188). Other parasitic infections such as Chagas disease, neurocysticercosis, schistosomiasis, and strongyloidiasis are exceedingly less common (189). Nocardiosis is usually observed between one and six months posttransplantation. The clinical presentation of nocardiosis includes pneumonia, CNS focal lesions, and cutaneous involvement (190–193). Brain abscesses due to Nocardia are multiple in up to 40% of the cases and may demonstrate ring enhancement. Diagnosis may be reached by direct observation of biological samples using modified Ziehl-Neelsen staining or Gram stain.

Infection in Organ Transplant Patients in the Critical Care Unit

477

BSI, Catheter-Related Infections, and Infective Endocarditis As other patients requiring intensive care, catheter-related bloodstream infections (CRBSI) are a potential threat for severe infection after SOT. In a recent study performed by our group in heart transplant recipients, CRBSI accounted for 16% of BSI in this population (194). In heart transplant recipients the incidence of bloodstream infection is 15.8%. Bloodstream infection (BSI) episodes were detected a median of 51 days after transplantation. The main BSI origins were lower respiratory tract (23%), urinary tract (20%), and catheter-related-BSI (16%). Gram-negative organisms predominated (55.3%), followed by gram-positive (44.6%). We found a clear relationship between time of onset and some characteristics of the BSI. During the first month after transplantation, 95% of the BSI were nosocomially acquired, and the main origins were intravenous (IV) catheter (32%), surgical site, and lower respiratory tract (LRT) (18% each). From month 2 to month 6, 70% of the BSI were nosocomially acquired, and the main origins were urinary tract infection (UTI) and LRT (25% each). After the sixth month, only 22% of the BSI episodes were nosocomial, and the most common portals of entry were LRT (33%), primary bacteremia (22%), and urinary tract infection UTI (17%) (p ¼ 0.1). Mortality was 59.2%, with 12.2% directly attributable to BSI. Independent risk factors for BSI after HT were hemodialysis (OR 6.5; 95% CI 3.2–13), prolonged ICU stay (OR 3.6; 95% CI 1.6–8.1), and viral infection (OR 2.1; 95% CI 1.1–4). BSI was a risk factor for mortality (OR 1.8; 95% CI 1.2–2.8) (194). CRBSI caused 15% of the febrile episodes of liver transplant recipients in the ICU (9). Although only 37% of the bacterial infections after liver transplantation occur more than 100 days after transplant, 60% of the cases of primary bacteremia after liver transplantation occur late (195). The incidence of BSI after OLT is 0.28 episodes/patient. BSI accounted for 36% of all major infections. Intravascular catheters were the most frequent source, and methicillin-resistant S. aureus was the most frequent pathogen causing bloodstream infections. In recent years a shift toward a higher importance of gram-negative microorganisms causing bacteremia has been observed (194,196). Gram-negative CRBSI, mainly if more than one case is detected, should always prompt exclusion of a nosocomial hazard, such as contamination of the infusate or transmission by the health-care workers (197,198). Seventy percent of the catheter-related and all bacteremias due to intraabdominal infections occurred 90 days, whereas 75% of the bacteremias due to biliary source occurred > 90 days after transplantation. Length of initial posttransplant ICU stay (p ¼ 0.014) and readmission to the ICU (p ¼ 0.003) were independently significant predictors of bloodstream infections. Forty percent of the candidemias occurred within 30 days of transplantation and were of unknown portal, whereas the portal in all candidemias occurring > 30 days posttransplant was known (catheter, hepatic abscess, and urinary tract). Mortality in patients with bloodstream infections was 52% (15/29) versus 9% (9/101) in patients without bloodstream infections (p ¼ 0.0001). In conclusion, intravascular catheters (and not intra-abdominal infections) have emerged as the most common source of BSI after OLT (199). In another study, primary (catheter-related) bacteremia (31%; 9 of 29 patients), pneumonia (24%; 7 of 29 patients), abdominal and/or biliary infections (14%; 4 of 29 patients), and wound infections (10%; 3 of 29 patients) were the predominant sources of bacteremia (200). Most important risk factors for CRBSI is the length of catheterization. Most catheters used in critically ill SOT patients are short-termed. They include central

478

Mun˜oz et al.

venous catheters, temporary hemodialysis catheters, peripheral venous catheters, and arterial cannulas. The site of central venous catheterization (internal jugular vein vs. the subclavian vein) does not seem to have an impact on the incidence of related infections as long as catheterization is performed by experienced personnel (201). S. aureus nasal carriage is associated with a higher risk of bacteremia (54); active surveillance cultures to detect colonization and implementation of targeted infection control interventions have proved to be effective in curtailing new acquisition of S. aureus colonization and in decreasing the rate of S. aureus infection in this population (202). Strict adherence to prophylactic guidelines may help reduce the incidence of these infections. Infective endocarditis is a rare event in SOT population (1.7–6%), but it may be an underappreciated sequela of hospital-acquired infection in transplant patients (50). The spectrum of organisms causing infective endocarditis was clearly different in transplant recipients than in the general population; 50% of the infections were due to Aspergillus fumigatus or S. aureus, but only 4% were due to viridans streptococci. Fungal infections predominated early (accounting for 6 of 10 cases of endocarditis within 30 days of transplantation), while bacterial infections caused most cases (80%) after this time. In 80% (37) of the 46 cases in transplant recipients, there was no underlying valvular disease. Seventy-four percent (34) of the 46 cases were associated with previous hospital-acquired infection, notably venous access device and wound infections. Three patients with S. aureus endocarditis had had an episode of S. aureus bacteremia more than three weeks prior to the diagnosis of endocarditis and had received treatment for the initial bacteremia of less than the duration of 14 days. The overall mortality rate was 57% (26 of 46 patients died), with 58% (15) of the 26 fatal cases not being suspected during life (50). CMV, toxoplasma, and parvovirus B19 may cause myocarditis in this population. Therapy of established infections is similar to that of other immunosuppressed patients. Fever of Unknown Origin Undoubtedly, the most common alarm sign suggesting infection is fever. In transplant recipients, fever has been defined as an oral temperature of 37.8 C on at least two occasions during a 24-hour period (9). Antimetabolite immunosuppressive drugs, MMF and azathioprine, are associated with significantly lower maximum temperatures and leukocyte counts (203). However, it is important to remember that fever and infections do not always come together. The absence of fever does not exclude infection. In fact, 40% of the liver recipients with documented infection (mainly fungal) were afebrile in a recent series (35). Absence of febrile response has been found to be a predictor of poor outcome in liver transplant recipients with bacteremia (200). In that series, the independent factors predictive of greater mortality were ICU stay at the time of bacteremia (100% vs. 47%; P ¼ 0.005), absence of chills (0% vs. 53%; P ¼ 0.005), lower temperature at the onset of bacteremia (99.2 F vs. 101.5 F; P ¼ 0.009), lower maximum temperature during the course of bacteremia (99.3 F vs. 102 F, P ¼ 0.008), greater serum bilirubin level (7.6 vs. 1.5 mg/dL; P ¼ 0.024), presence of abnormal blood pressure (80% vs. 16%; P ¼ 0.0013), and greater prothrombin time (15.6 seconds vs. 13.3 seconds; P ¼ 0.013). A major difference with immunocompetent critical patients is that the list of potential etiological agents is much longer and is influenced by time elapsed from transplantation. CMV (as main offender or as copathogen) should be considered in practically all-infectious complications in this population. Accordingly, a sample

Infection in Organ Transplant Patients in the Critical Care Unit

479

for CMV antigenemia (or PCR if available) should always be obtained. Other viruses such as adenovirus, influenza A, or HHV-6 may also cause severe infections after SOT and can be recovered from respiratory samples or blood. If indicated, invasive diagnostic procedures should be performed rapidly and a serum sample stored. Bacterial infections must always be considered and urine and blood cultures obtained before starting therapy. Diagnosis of catheter-related infections without removing the devices may be attempted in stable patients. Lysis centrifugation blood cultures and hub and skin cultures have a high negative predictive value (204). The first steps for diagnosis of pneumonia should include a chest X ray and culture of expectorated sputum or bronchoaspirate (submitted for virus, bacteria, mycobacteria, and fungus). A CT scan or ultrasonography may also be ordered to exclude the presence of collections in the proximity of the surgical area. Lumbar puncture and cranial CT (including the paranasal sinus) must be performed if neurological symptoms or signs are detected. In case of diarrhea, C. difficile should be investigated. Cultures and PCR for detection of M. tuberculosis should be ordered for all transplant recipients with suspicion of infection. Fungal infections should be aggressively pursued in colonized patients and in patients with risk factors. Early stages of fungal infection may be very difficult to detect (95,205). Isolation of Candida or Aspergillus from superficial sites may indicate infection. Fundi examination, blood and respiratory cultures and Aspergillus and Cryptococcus antigen detection tests must be performed. Parasitic infections are uncommon, but toxoplasmosis and leishmaniasis should be considered if diagnosis remains elusive. Serology or bone marrow cultures usually provide the diagnosis. The possibility of a Toxoplasma primary infection should be considered when a seronegative recipient receives an allograft from a seropositive donor. HT recipients are more susceptible to toxoplasmosis, which may be transmitted with the allograft and occasionally requires ICU admission. The risk of primary toxoplasmosis (R-Dþ) is over 50% in HT, 20% after liver transplantation, and < 1% after kidney transplantation. Patients with toxoplasmosis have fever, altered mental status, focal neurological signs, myalgias, myocarditis, and lung infiltrates. Allograft-transmitted toxoplasmosis is more often associated with acute disease (61%) than with reactivation of latent infection (7%). Lethal cases associated to hemophagocytic syndrome have been described (206). Leishmaniasis is another parasitic infection that should be excluded, though it is exceedingly uncommon after SOT. It may present as fever, pancytopenia, and splenomegaly. Multimodality imaging such the use of combined indium-labeled WBC scintigraphy and CT allowed the detection of infection within retained left ventricular assist device tubing in a heart transplant recipient with a diagnosis of fever of unknown origin (207). Noninfectious Causes of Fever Both infectious and noninfectious causes of fever should be considered when approaching a febrile SOT patient. In a recent series, 87% of the febrile episodes detected in OLT in the ICU were due to infections, and 13% were noninfectious (9). Rejection, malignancy, adrenal insufficiency, and drug fever were the most common noninfectious causes. Fever is common in the first 48 hours after surgery and after certain procedures. If it is not persistent or accompanied by other signs or symptoms it should not trigger any diagnostic action. Acute rejection accounts for 4% to 17% of

480

Mun˜oz et al.

the noninfectious febrile episodes (208). It is usually related to an impairment of the allograft function and requires histological confirmation. It is more common in the first six months, especially in the first 16 days after transplantation in one study (209). It is important to remember that severe graft rejection and increased immunosuppression could stimulate cooperatively active CMV (210,211). Malignancy, mainly lymphoproliferative disease, is relatively common after SOT and may initially present as a febrile episode (80%) (212). It usually occurs longer after transplantation (208). Acute adrenal insufficiency should be excluded in SOT patients admitted to an ICU because of sepsis or surgery, mainly when corticosteroids have been withdrawn and drugs that accelerate the degradation of cortisol (phenytoin and rifampin) are administered (213). However, although analytical adrenal insufficiency is frequent in SOT patients, prospective studies suggest that supplemental steroids are not needed in most cases even under stress (214–216). Another setting of potential adrenal insufficiency is renal transplants that return to dialysis (217,218). Occasionally, lymphoproliferative disease may present with adrenal insufficiency after liver transplantation (219). Drugs such as OKT3, antithymocyte globulin (ATG), everolimus, antimicrobials, interferon, anticonvulsants, etc. may also cause fever in this population (220). The temporal relationship with the drug is usually a diagnostic clue. New induction therapies such as basiliximab are related to fewer side effects and fewer CMV infections (221). Other causes of noninfectious fever include thromboembolic disease, hematoma reabsorption, pericardial effusions, tissue infarction, hemolytic uremic syndrome, and transfusion reaction. Noncardiogenic pulmonary edema (pulmonary reimplantation response) is a common finding after lung transplantation (50–60%) and may occasionally lead to a differential diagnosis with pneumonia. It gives rise to prolonged mechanical ventilation and ICU stay but does not affect survival (222).

MANAGEMENT Diagnostic Approach As we mentioned before, the diagnostic approach to a critically ill SOT with suspected infection should take into account the time onwards from transplantation (Table 1) and previous complications such as episodes of rejection, surgical or technical problems, reactivation of a latent infection, etc The findings provided by the anamnesis and physical examination (see previous parts of this chapter) may suggest a focus causative of the fever (pneumonia, wound infection, etc.). In this situation, a list of possible pathogens as well as necessary samples and tests for diagnosis should be elaborated. In most cases, analytical and imaging studies will also be ordered. Samples for culture should be obtained before starting empirical antimicrobial therapy. In a recent study, 79% of the infections associated with fever in the liver recipients in the ICU were bacterial, 9% viral, and 9% fungal. Accordingly, blood cultures are practically always needed. Bacteremia is present in 45% of the febrile critical SOT patients, and its origin must always be investigated. In liver recipients the most common sources are IV devices, lung, biliary tree, and wound infections. Accordingly, the entry site of the catheters must be examined. MRSA and P. aeruginosa caused 65% of the bacteremias in ICU patients (7). Lack of febrile response in bacteremic OLT recipients portended a poorer outcome (195).

Infection in Organ Transplant Patients in the Critical Care Unit

481

In heart transplant recipients, the main BSI origins were lower respiratory tract, urinary tract, and CRBSI, which should always be investigated (194). If focal signs of infections are present, appropriate samples must be sent to the laboratory (catheter tips, wound exudate, CSF, etc.) as in any other critical patient. When a collection of fluid or pus is to be sampled, aspirated material provides more valuable information. Length of stay in the ICU is also a determinant factor, which may help find the origin of the infection. Pneumonia is more common in the first seven days of ICU stay, while CRI incidence tripled after the first week. Information on some of the most severe infections may be obtained rapidly when the clinician and the microbiology laboratory communicate effectively and the best specimen type and test are selected. Antigen detection tests for adenovirus, HSV, Influenza A, respiratory syncytical virus (RSV), rotavirus etc., are available. Most common herpesviruses can be easily cultured and detected. Gram stain requires expertise but may provide valuable rapid information (five minutes) on the quality of the specimen and whether gram-negative or positive rods or cocci are present. It may reveal yeasts and occasionally molds, parasites, Nocardia, and even mycobacteria. The amount of material and the number of organisms limit detection sensitivity. Continuous agitation blood cultures have significantly reduced the detection time to less than 24 hours for bacterial isolates. Direct testing of specimens with antigen assays are mainly used for CSF samples (Neisseria meningitidis, S. pneumoniae, and C. neoformans). Group A streptococci, C. difficile, and C. trachomatis antigen detection tests are also available. Specific stains for Legionella direct fluorescence assay (DFA) and Bordetella pertussis are offered by most laboratories. Legionella urinary antigen test will be very useful in pneumonias caused by L. pneumophila serotype 1, and S. pneumoniae antigenuria can also be rapidly investigated. HIV infection, Brucella, and syphilis are some of the infections that can be rapidly diagnosed serologically. Acid-fast stain and fluorochrome stains for mycobacteria or Nocardia require a more prolonged laboratory procedure (30–60 minutes). New techniques, such as PCR and quantification of interferon-gamma, have been developed to achieve more rapid and accurate diagnoses. M. tuberculosis complex PCR is very effective in smear-positive specimens. In smear-negative samples sensitivity is 70% (74). Fungal elements may be rapidly detected in wet mounts with potassium hydroxide or immunofluorescent Calcofluor white stain. An India ink preparation allows the identification of encapsulated C. neoformans, particularly in CSF in approximately 50% of patients. The latex agglutination test or enzyme immunoassay (EIA) cryptococcal antigen have greater sensitivity. Fluorescent antibody stains or toluidine blue O permits the detection of P. jiroveci. Antigen detection for Histoplasma capsulatum is quite sensitive, and the detection of Aspergillus antigen is useful, although its efficiency is lower than in hematological patients (223–225). Management Fever is not harmful by itself, and accordingly it should not be systematically eliminated. In fact, it has been demonstrated that fever enhances several host defense mechanisms (chemotaxis, phagocytosis, and opsonization) (119). Besides, antibiotics may be more active at higher body temperatures. If provided, antipyretic drugs should be administered at regular intervals to avoid recurrent shivering and an associated increase in metabolic demand.

482

Mun˜oz et al.

After obtaining the previously mentioned samples, empiric antibiotics should be promptly started in all transplant patients with suspicion of infection and toxic or unstable situation. They are also recommended if a focus of infection is apparent, in the early posttransplant setting in which nosocomial infection is very common, or when there has been a recent increase of immunosuppression. In a stable patient without a clear source of infection further diagnostic testing should carried out and noninfectious causes considered. We have recently demonstrated that only 58.5% of patients with BSI received appropriate empirical antimicrobial therapy. Inadequate treatment was related to a longer hospital stay, a higher mean risk of CDAD, a higher mean overall mortality rate, and a higher risk of infection-related mortality (226). So once blood cultures are obtained, empirical broad-spectrum antimicrobials guided by the clinical condition of the patient and the presumed origin should be promptly started. When results of blood cultures are available, antibiotics should be adjusted according to susceptibility patterns of the isolates. This antibacterial de-escalation strategy attempts to balance the need to provide appropriate, initial antibacterial treatment while limiting the emergence of antibacterial resistance. The selection of the antimicrobial should be based on the likely origin of the infection, prevalent bacterial flora, rate of antimicrobial resistance, and previous use of antimicrobials by the patient. In our series of bacteremia in HT recipients gram-negative microorganisms predominated (55.3%), followed by gram-positive microorganisms (44.6%). Gram-negatives accounted for 54% of infections in the first month, 50% during months 2 to 6, and 72% of infections occurring afterwards ( p ¼ 0.3) (194). The possibility of drug interactions mainly with cyclosporine and tacrolimus is very real and impacts significantly on the choice of antimicrobial. There are three categories of antimicrobial interaction with cyclosporine and tacrolimus. First, the antimicrobial agent (e.g., rifampin, isoniazid, and nafcillin) upregulates the metabolism of the immunosuppressive drugs, resulting in decreased blood levels and an increased possibility of allograft rejection. Second, the antimicrobial agent (e.g., the macrolides erythromycin, clarithromycin, and to a lesser extent azithromycin, or the azoles ketoconazole, itraconazole, and to a lesser extent fluconazole) downregulates the metabolism of the immunosuppressive drugs, resulting in increased blood levels and an increased possibility of nephrotoxicity and overimmunosuppression. And last, there may be synergistic nephrotoxicity, when therapeutic levels of the immunosuppressive agents are combined with therapeutic levels of aminoglycosides, amphotericin, and vancomycin, and high therapeutic doses of trimethoprim-sulfamethoxazole and fluoroquinolones. Outcome of Febrile Processes of SOT Recipients in the ICU SOT patients have higher risk of dying after an ICU admission than the general population, and in most series it is a poor prognostic factor (227,228). However, the overall prognosis is better than that of bone marrow recipients (229–231). The overall ICU mortality of SOT patients was 18% in a recent series, and infection was the major cause of death (disseminated mycoses, hepatitis C virus (HCV), multiorganic failure, hepatic artery thrombosis with sepsis, and primary nonfunction of the graft). Mortality of febrile liver recipients at 14 days (24% vs. 0%, p ¼ 0.001) and at 30 days (34% vs. 5%, p ¼ 0.001) was significantly higher in the ICU, as compared to

Infection in Organ Transplant Patients in the Critical Care Unit

483

non-ICU patients (9). Mortality of OLT with lung infiltrates in the ICU was 28%. Pneumonia, creatinine level > 1.5 mg/dL, higher blood urea nitrogen, and worse acute physiology and chronic health evaluation (APACHE) neurological score were predictors of poor outcome (35). The need for mechanical ventilation was an independently significant predictor of mortality (7). Infection was a risk factor for early renal dysfunction (232). Need for preoperative ICU care was predictive of an increased risk of death in OLT patients waiting for retransplantation (228). Infection is also a leading cause of death in heart recipients (30% of early deaths, 45% of deaths from one to three months, and 9.7% thereafter) (233). Overall, 31% of the patients with pneumonia died (Aspergillus 62%; CMV 13%; nosocomial bacteria 26%). Mortality was 100% in patients requiring mechanical ventilation (7 out of 13 Aspergillus, 5 out of 11 P. carinii, 1 out of 8 CMV) (58). From 51 lung transplant recipients who required admission to the ICU at the Duke University Medical Center, 53% required mechanical ventilation, and 37% died (59% of those requiring mechanical ventilation) (234). In other series, mortality of lung transplant recipients requiring admission to a medical intensive care unit (MICU) was 37%. A preadmission diagnosis of bronchiolitis obliterans syndrome, APACHE III scores, nonpulmonary organ system dysfunction, initial serum albumin level, and duration of mechanical ventilation are important prognostic factors (27). Mortality of renal transplant recipients in the ICU was 11% in a recent series, and infection caused six out of seven deaths (235). PREVENTION Organ transplant patients admitted to the ICU should receive all measures available to prevent nosocomial infection. The first one could be ‘‘avoid the admission to the unit itself,’’ which has been demonstrated to be a very stress-inducing situation for transplant recipients (236). In one recent study it was determined the proportion of liver transplant patients who could be extubated immediately after surgery and transferred to the surgical ward without intervening ICU care. Of 147 patients, 36 patients did not meet postsurgical criteria for early extubation, and 111 patients were successfully extubated. Eighty-three extubated patients were transferred to the surgical ward after a routine admission to the postoperative care unit. Only three patients who were transferred to the surgical ward experienced complications that required a greater intensity of nursing care. A learning curve detected during the three-year study period showed that attempts to extubate increased from 73% to 96%, and triage to the surgical ward increased from 52% to 82% without compromising patient safety. The protocol resulted in a one-day reduction in ICU use in 75.5% of study subjects (237). The same approach can be extended to the use of IV catheters or indwelling bladder catheters, which should be withdrawn as soon as possible. Other measures such as selective gastrointestinal decontamination (238), use of gowns, or high efficiency particulate air filters (HEPA) filters have not demonstrated so clearly an impact on the reduction of mortality or even nosocomial infections. REFERENCES 1. Miller LW, Naftel DC, Bourge RC, et al. Infection after heart transplantation: a multiinstitutional study. Cardiac Transplant Research Database Group. J Heart Lung Transplant 1994; 13(3):381–392.

484

Mun˜oz et al.

2. Plo¨chl W, Pezawas L, Artemiou O, Grimm M, Klepetko W, Hiesmayr M. Nutritional status, ICU duration and ICU mortality in lung transplant recipients. Intens Care Med 1996; 22(11):1179–1185. 3. Hsu J, Griffith BP, Dowling RD, et al. Infections in mortally ill cardiac transplant recipients. J Thorac Cardiovasc Surg 1989, 98:506–509. 4. Cisneros Alonso C, Montero Castillo A, Moreno Gonza´lez E, Garcı´a Garcı´a I, Guille´n Ramı´rez F, Garcı´a Fuentes C. Complications of liver transplant in intensive care. Experience in 130 cases. Rev Clin Esp 1991; 189(6):264–267. 5. Plevak DJ, Southorn PA, Narr BJ. Intensive-care unit experience in the Mayo liver transplantation program: the first 100 cases. Mayo Clin Proc 1989; 64:433–445. 6. Bindi ML, Biancofiore G, Pasquini C, et al. Pancreas transplantation: problems and prospects in intensive care units. Minerva Anestesiol 2005; 71(5):207–221. 7. Singh N, Gayowski T, Wagener MM. Intensive care unit management in liver transplant recipients: beneficial effect on survival and preservation of quality of life. Clin Transpl 1997; 11(2):113–120. 8. Trzeciak S, Sharer R, Piper D, et al. Infections and severe sepsis in solid-organ transplant patients admitted from a university-based ED. Am J Emerg Med 2004; 22(7): 530–533. 9. Singh N, Chang FY, Gayowski T, Wagener M, Marino IR. Fever in liver transplant recipients in the intensive care unit. Clin Transpl 1999; 13(6):504–511. 10. Viscoli C, Dimitri P, Di Domenico S, Mannelli S, Dodi F, Veroni L. Infectious complications in liver transplant in Italy: current status and prospectives. Recenti Prog Med 2001; 92(1):16–31. 11. Montoya JG, Giraldo LF, Efron B, et al. Infectious complications among 620 consecutive heart transplant patients at Stanford University Medical Center. Clin Infect Dis 2001; 33(5):629–640. 12. Rubin RH. The prevention and treatment of infectious disease in the transplant patient: where are we now and where do we need to go? Transpl Infect Dis 2004; 6(1):1–2. 13. Rubin RH, Marty FM. Principles of antimicrobial therapy in the transplant patient. Transpl Infect Dis 2004; 6(3):97–100. 14. Martinelli L, Rinaldi M, Pederzolli C, et al. Different results of cardiac transplantation in patients with ischemic and dilated cardiomyopathy. Eur J Cardiothorac Surg 1995; 9(11):644–650. 15. Mun˜oz P, Mun˜oz RM, Palomo J, Rodrı´guez Creixe´ms M, Mun˜oz R, Bouza E. Pneumocystis carinii infections in heart transplant patients. Twice a week prophylaxis. Medicine (Baltimore) 1997; 76:415–422. 16. Edwards LB, Keck BM. Thoracic organ transplantation in the US. Clin Transpl 2002:29–40. 17. Detre KM, Belle SH, Carr MA, et al. A report from the NIDDK Liver Transplantation Database. Clin Transpl 1989:129–141. 18. Ben Hamida C, Lauzet JY, Rezaiguia-Delclaux S, et al. Effect of severe thrombocytopenia on patient outcome after liver transplantation. Intens Care Med 2003; 29(5): 756–762. 19. Aggarwal A, Ong JP, Goormastic M, et al. Survival and resource utilization in liver transplant recipients: the impact of admission to the intensive care unit. Transplant Proc 2003; 35(8):2998–3002. 20. Wiebe K, Wahlers T, Harringer W, vd Hardt H, Fabel H, Haverich A. Lung transplantation for cystic fibrosis—a single center experience over 8 years. Eur J Cardiothorac Surg 1998; 14(2):191–196. 21. Madden BP, Kamalvand K, Chan CM, Khaghani A, Hodson ME, Yacoub M. The medical management of patients with cystic fibrosis following heart–lung transplantation. Eur Respir J 1993; 6(7):965–970. 22. Nath DS, Kandaswamy R, Gruessner R, Sutherland DE, Dunn DL, Humar A. Fungal infections in transplant recipients receiving alemtuzumab. Transplant Proc 2005; 37(2): 934–936.

Infection in Organ Transplant Patients in the Critical Care Unit

485

23. Argenziano M, Catanese KA, Moazami N, et al. The influence of infection on survival and successful transplantation in patients with left ventricular assist devices. J Heart Lung Transplant 1997; 16(8):822–831. 24. Fischer SA, Trenholme GM, Costanzo MR, Piccione W. Infectious complications in left ventricular assist device recipients. Clin Infect Dis 1997; 24(1):18–23. 25. Masters RC, Hendry PJ, Davies RA, et al. Cardiac transplantation after mechanical circulatory support: a canadian perspective. Ann Thorac Surg 1996; 61:1734–1739. 26. Bouza E, Mun˜oz P, Rodriguez C, et al. Endotipsitis: an emerging prosthetic-related infection in patients with portal hypertension. Diagn Microbiol Infect Dis 2004; 49(2):77–82. 27. Pietrantoni C, Minai OA, Yu NC, et al. Respiratory failure and sepsis are the major causes of ICU admissions and mortality in survivors of lung transplants. Chest 2003; 123(2):504–509. 28. Lardinois D, Banysch M, Korom S, et al. Extended donor lungs: eleven years experience in a consecutive series. Eur J Cardiothorac Surg 2005; 27(5):762–767. 29. Cuende N, Miranda B, Canon JF, Garrido G, Matesanz R. Donor characteristics associated with liver graft survival. Transplantation 2005; 79(10):1445–1452. 30. Ostrowsky BE, Venkataraman L, EM DA, Gold HS, DeGirolami PC, Samore MH. Vancomycin-resistant enterococci in intensive care units: high frequency of stool carriage during a non-outbreak period. Arch Intern Med 1999; 159(13):1467–1472. 31. Mort TC, Yeston NS. The relationship of pre mortem diagnoses and post mortem findings in a surgical intensive care unit [see comments]. Crit Care Med 1999; 27(2):299–303. 32. Singh N, Paterson DL. Mycobacterium tuberculosis infection in solid-organ transplant recipients: impact and implications for management. Clin Infect Dis 1998; 27(5):1266– 1277. 33. Mun˜oz P, Guinea J, Pelaez T, Duran C, Blanco JL, Bouza E. Nosocomial invasive aspergillosis in a heart transplant patient acquired during a break in the HEPA air filtration system. Transpl Infect Dis 2004; 6(1):50–54. 34. Mun˜oz P, Arencibia J, Rodriguez C, et al. Trimethoprim-sulfamethoxazole as toxoplasmosis prophylaxis for heart transplant recipients. Clin Infect Dis 2003; 36(7):932–933 (author reply 3). 35. Singh N, Gayowski T, Wagener MM, Marino IR. Pulmonary infiltrates in liver transplant recipients in the intensive care unit. Transplantation 1999; 67(8):1138–1144. 36. Papanicolaou GA, Meyers BR, Meyers J, et al. Nosocomial infections with vancomycinresistant Enterococcus faecium in liver transplant recipients: risk factors for acquisition and mortality. Clin Infect Dis 1996; 23(4):760–766. 37. Duchini A, Goss JA, Karpen S, Pockros PJ. Vaccinations for adult solid-organ transplant recipients: current recommendations and protocols. Clin Microbiol Rev 2003; 16(3):357–364. 38. Braddy CM, Heilman RL, Blair JE. Coccidioidomycosis after renal transplantation in an endemic area. Am J Transplant 2006; 6(2):340–345. 39. Martı´n-Rabada´n P, Mun˜oz P, Palomo J, Bouza E. Strongyloidiasis: the Harada-Mori test revisited. Clin Microbiol Infection 1999; 5:374–376. 40. Tan HP, Stephen Dumler J, Maley WR, et al. Human monocytic ehrlichiosis: an emerging pathogen in transplantation. Transplantation 2001; 71(11):1678–1680. 41. Lafayette RA, Pare´ G, Schmid CH, King AJ, Rohrer RJ, Nasraway SA. Pretransplant renal dysfunction predicts poorer outcome in liver transplantation. Clin Nephrol 1997; 48(3):159–164. 42. Descheˆnes M, Belle SH, Krom RA, Zetterman RK, Lake JR. Early allograft dysfunction after liver transplantation: a definition and predictors of outcome. National Institute of Diabetes and Digestive and Kidney Diseases Liver Transplantation Database. Transplantation 1998; 66(3):302–310. 43. Reilly J, Mehta R, Teperman L, et al. Nutritional support after liver transplantation: a randomized prospective study [see comments]. JPEN 1990; 14(4):386–391.

486

Mun˜oz et al.

44. Gurakar A, Hassanein T, Van Thiel DH. Right diaphragmatic paralysis following orthotopic liver transplantation. J Okla State Med Assoc 1995; 88(4):149–153. 45. Paterson DL, Staplefeldt WH, Wagener MM, Gayowski T, Marino IR, Singh N. Intraoperative hypothermia is an independent risk factor for early cytomegalovirus infection in liver transplant recipients. Transplantation 1999; 67(8):1151–1155. 46. Shorr AF, Jackson WL. Transfusion practice and nosocomial infection: assessing the evidence. Curr Opin Crit Care 2005; 11(5):468–472. 47. Parker BM, Irefin SA, Sabharwal V, et al. Leukocyte reduction during orthotopic liver transplantation and postoperative outcome: a pilot study. J Clin Anesth 2004; 16(1):18–24. 48. Naka T, Wan L, Bellomo R, et al. Kidney failure associated with liver transplantation or liver failure: the impact of continuous veno-venous hemofiltration. Int J Artif Organs 2004; 27(11):949–955. 49. Maroto LC, Aguado JM, Carrascal Y, et al. Role of epicardial pacing wire cultures in the diagnosis of poststernotomy mediastinitis. Clin Infect Dis 1997; 24(3):419–421. 50. Paterson DL, Dominguez EA, Chang FY, Snydman DR, Singh N. Infective endocarditis in solid organ transplant recipients. Clin Infect Dis 1998; 26(3):689–694. 51. Mugnani G, Bergami M, Lazzarotto T, Bedani PL. Intestinal infection by cytomegalovirus in kidney transplantation: diagnostic difficulty in the course of mycophenolate mofetil therapy. G Ital Nefrol 2002; 19(4):483–484. 52. Bouza E, Merino P, Mun˜oz P, Sa´nchez-Carrillo C, Ya´n˜ez J, Corte´s C. Ocular tuberculosis: a prospective study in a General Hospital. Medicine (Baltimore) 1997; 76:53–61. 53. Bouza E, Cobo-Soriano R, Rodrı´guez-Cre´ixems M, Mun˜oz P, Sua´rez-Leoz M, Corte´s C. A prospective search for ocular lesions in hospitalized patients with significant bacteremia. Clin Infect Dis 2000; 30:306–312. 54. Chang FY, Singh N, Gayowski T, Drenning SD, Wagener MM, Marino IR. Staphylococcus aureus nasal colonization association with infections in liver transplant recipients. Transplanation 1998; 65(9):1169–1172. 55. Mermel LA, Maki DG. Bacterial pneumonia in solid organ transplantation. Semin Respir Infect 1990; 5(1):10–29. 56. Jensen WA, Rose RM, Hammer SM, et al. Pulmonary complications of orthotopic liver transplantation. Transplantation 1986; 42(6):484. 57. Jimenez-Jambrina M, Hernandez A, Cordero E, et al. Pneumonia after heart transplantation in the XXI century: a multicenter prospective study. In: 45th Interscience Conference on Antimicrobial Agents and Chemotherapy, 2005 (K-1561/370). 58. Cisneros JM, Mun˜oz P, Torre-Cisneros J, et al. Pneumonia after heart transplantation: a multiinstitutional study. Clin Infect Dis 1998; 27:324–331. 59. Fraser TG, Zembower TR, Lynch P, et al. Cavitary Legionella pneumonia in a liver transplant recipient. Transpl Infect Dis 2004; 6(2):77–80. 60. Singh N, Gayowski T, Wagener M, Marino IR, Yu VL. Pulmonary infections in liver transplant recipients receiving tacrolimus. Changing pattern of microbial etiologies. Transplantation 1996; 61(3):396–401. 61. Nichols L, Strollo DC, Kusne S. Legionellosis in a lung transplant recipient obscured by cytomegalovirus infection and Clostridium difficile colitis. Transpl Infect Dis 2002; 4(1):41–45. 62. Horbach I, Fehrenbach FJ. Legionellosis in heart transplant recipients. Infection 1990; 18(6):361–363. 63. Gupta RK, Jain M, Garg R. Pneumocystis carinii pneumonia after renal transplantation. Indian J Pathol Microbiol 2004; 47(4):474–476. 64. Renoult E, Georges E, Biava MF, et al. Toxoplasmosis in kidney transplant recipients: report of six cases and review. Clin Infect Dis 1997; 24(4):625–634. 65. Chang GC, Wu CL, Pan SH, et al. The diagnosis of pneumonia in renal transplant recipients using invasive and noninvasive procedures. Chest 2004; 125(2):541–547. 66. Ampel NM, Wing EJ. Legionella infection in transplant patients. Semin Respir Infect 1990; 5(1):30–37.

Infection in Organ Transplant Patients in the Critical Care Unit

487

67. Miller R, Burton NA, Karwande SV, Jones KW, Doty DB, Gay WA Jr. Early, aggressive open lung biopsy in heart transplant recipients. J Heart Transplant 1987; 6(2): 96–99. 68. Dowling JN, Pasculle AW, Frola FN, Zaphyr MK, Yee RB. Infections caused by Legionella micdadei and Legionella pneumophila among renal transplant recipients. J Infect Dis 1984; 149(5):703–713. 69. Singh N, Muder RR, Yu VL, Gayowski T. Legionella infection in liver transplant recipients: implications for management. Transplantation 1993; 56(6):1549–1551. 70. Ogunc G, Ozdemir T, Vural T, Suleymanlar G, Akaydin M, Karpuzoglu T. Legionnaires’ disease following kidney transplantation. Mikrobiyol Bul 1993; 27(2):137–142. 71. La Scola B, Michel G, Raoult D. Isolation of Legionella pneumophila by centrifugation of shell vial cell cultures from multiple liver and lung abscesses. J Clin Microbiol 1999; 37(3):785–787. 72. Chaberny IF, Ziesing S, Gastmeier P. Legionella prevention in intensive care units. Anasthesiol Intensivmed Notfallmed Schmerzther 2004; 39(3):127–131. 73. Vonberg RP, Eckmanns T, Bruderek J, Ruden H, Gastmeier P. Use of terminal tap water filter systems for prevention of nosocomial legionellosis. J Hosp Infect 2005; 60(2):159–162. 74. Mun˜oz P, Rodriguez C, Bouza E. Mycobacterium tuberculosis infection in recipients of solid organ transplants. Clin Infect Dis 2005; 40(4):581–587. 75. Mun˜oz P, Palomo J, Mun˜oz R, Rodrı´guez-Creixe´ms M, Pelaez T, Bouza E. Tuberculosis in heart transplant recipients. Clin Infect Dis 1995; 21:398–402. 76. Paterson DL, Singh N, Gayowski T, Marino IR. Pulmonary nodules in liver transplant recipients. Medicine 1998; 77(1):50–58. 77. Aguado JM, Herrero JA, Gavalda J, et al. Clinical presentation and outcome of tuberculosis in kidney, liver, and heart transplant recipients in Spain. Spanish Transplantation Infection Study Group, GESITRA. Transplantation 1997; 63(9):1278–1286. 78. Mun˜oz P, Burillo A, Palomo J, Rodrı´guez-Creixe´ms M, Bouza E. Rhodococcus equi infection in transplant recipients: casereview of the literature. Transplantation 1997; 65:449–453. 79. Wiesmayr S, Stelzmueller I, Tabarelli W, et al. Nocardiosis following solid organ transplantation: a single-centre experience. Transpl Int 2005; 18(9):1048–1053. 80. Vigano SM, Edefonti A, Ferraresso M, et al. Successful medical treatment of multiple brain abscesses due to Nocardia farcinica in a paediatric renal transplant recipient. Pediatr Nephrol 2005; 20(8):1186–1188. 81. Queipo-Zaragoza JA, Broseta-Rico E, Alapont-Alacreu JM, Santos-Durantez M, Sanchez-Plumed J, Jimenez-Cruz JF. Nocardial infection in immunosuppressed kidney transplant recipients. Scand J Urol Nephrol 2004; 38(2):168–173. 82. Kahraman S, Genctoy G, Arici M, Cetinkaya Y, Altun B, Caglar S. Septic arthritis caused by Nocardia asteroides in a renal transplant recipient. Transplant Proc 2004; 36(5):1415–1418. 83. Nouza M, Prat V. Lung infection after kidney transplantation. I. Etiology, pathogenesis and clinical picture. Cas Lek Cesk 1990; 129(21):641–644. 84. Mun˜oz P, Burillo A, Bouza E. Mold infection in solid organ transplant patients. In: Bowden RA, Ljungman P, Paya C, eds. Transplant Infections. 2nd ed. Lippincott: Williams & Wilkins, 2003. 85. Gavalda` J, Len O, Rovira M, et al. Epidemiology of invasive fungal infections (IFI) in solid organ (SOT) and hematopoietic stem cell (HSCT) transplant recipients: a prospective study from Resitra. In: 45th Interscience Conference on Antimicrobial Agents and Chemotherapy, 2005 (M-990/461). 86. Dummer JS, Lazariashvilli N, Barnes J, Ninan M, Milstone AP. A survey of anti-fungal management in lung transplantation. J Heart Lung Transplant 2004; 23(12):1376–1381. 87. Monforte V, Roman A, Gavalda J, et al. Nebulized amphotericin B prophylaxis for Aspergillus infection in lung transplantation: study of risk factors. J Heart Lung Transplant 2001; 20(12):1274–1281.

488

Mun˜oz et al.

88. Kramer MR, Denning DW, Marshall SE, et al. Ulcerative tracheobronchitis after lung transplantation. A new form of invasive aspergillosis. Am Rev Respir Dis 1991; 144(3 Pt 1):552–556. 89. Mun˜oz P, Palomo J, Guembe P, Rodriguez-Creixems M, Gijon P, Bouza E. Lung nodular lesions in heart transplant recipients. J Heart Lung Transplant 2000; 19(7): 660–667. 90. Mun˜oz P, Rodriguez C, Bouza E, et al. Risk factors of invasive aspergillosis after heart transplantation: protective role of oral itraconazole prophylaxis. Am J Transplant 2004; 4(4):636–643. 91. Singh N, Arnow PM, Bonham A, et al. Invasive aspergillosis in liver transplant recipients in the 1990s. Transplantation 1997; 64(5):716–720. 92. Singh N, Gayowski T, Wagener MM, Doyle H, Marino IR. Invasive fungal infections in liver transplant recipients receiving tacrolimus as the primary immunosuppressive agent. Clin Infect Dis 1997; 24(2):179–184. 93. Mun˜oz P, Burillo A, Bouza E. Environmental surveillance and other control measures in the prevention of nosocomial fungal infections. Clin Microbiol Infect 2001; 7(suppl 2):38–45. 94. Gavalda J, Len O, San Juan R, et al. Risk factors for invasive aspergillosis in solid-organ transplant recipients: a case-control study. Clin Infect Dis 2005; 41(1):52–59. 95. Paterson DL, Singh N. Invasive aspergillosis in transplant recipients. Medicine 1999; 78(2):123–138. 96. Husain S, Alexander BD, Mun˜oz P, et al. Opportunistic mycelial fungal infections in organ transplant recipients: emerging importance of non-Aspergillus mycelial fungi. Clin Infect Dis 2003; 37(2):221–229. 97. Husain S, Mun˜oz P, Forrest G, et al. Infections due to Scedosporium apiospermum and Scedosporium prolificans in transplant recipients: clinical characteristics and impact of antifungal agent therapy on outcome. Clin Infect Dis 2005; 40(1):89–99. 98. Singh N, Lortholary O, Alexander BD, et al. Antifungal management practices and evolution of infection in organ transplant recipients with Cryptococcus neoformans infection. Transplantation 2005; 80(8):1033–1039. 99. Singh N, Lortholary O, Alexander BD, et al. Allograft loss in renal transplant recipients with Cryptococcus neoformans associated immune reconstitution syndrome. Transplantation 2005; 80(8):1131–1133. 100. el-Ebiary M, Torres A, Fabregas N, et al. Significance of the isolation of Candida species from respiratory samples in critically ill, non-neutropenic patients. An immediate postmortem histologic study. Am J Respir Crit Care Med 1997; 156(2 Pt 1):583–590. 101. Park KY, Park CH. Candida infection in a stent inserted for tracheal stenosis after heart lung transplantation. Ann Thorac Surg 2005; 79(3):1054–1056. 102. Palmer SM, Perfect JR, Howell DN, et al. Candidal anastomotic infection in lung transplant recipients: successful treatment with a combination of systemic and inhaled antifungal agents. J Heart Lung Transplant 1998; 17(10):1029–1033. 103. Grossi P, Farina C, Fiocchi R, Dalla Gasperina D. Prevalence and outcome of invasive fungal infections in 1,963 thoracic organ transplant recipients: a multicenter retrospective study. Italian. Study Group of Fungal Infections in Thoracic Organ Transplant Recipients. Transplantation 2000; 70(1):112–116. 104. Horvath J, Dummer S, Loyd J, Walker B, Merrill WH, Frist WH. Infection in the transplanted and native lung after single lung transplantation. Chest 1993; 104(3): 681–685. 105. Sun Q, Li L, Ji S, et al. Variation of CD4þ and CD8þ T lymphocytes as predictor of outcome in renal allograft recipients who developed acute respiratory distress syndrome caused by cytomegalovirus pneumonia. Transplant Proc 2005; 37(5):2118–2121. 106. Liebau P, Kuse E, Winkler M, et al. Management of herpes simplex virus type 1 pneumonia following liver transplantation. Infection 1996; 24(2):130–135.

Infection in Organ Transplant Patients in the Critical Care Unit

489

107. Weiss RL, Colby TV, Spruance SL, Salmon VC, Hammond ME. Simultaneous cytomegalovirus and herpes simplex virus pneumonia. Arch Pathol Lab Med 1987; 111(3): 242–245. 108. Friedrichs N, Eis-Hubinger AM, Heim A, Platen E, Zhou H, Buettner R. Acute adenoviral infection of a graft by serotype 35 following renal transplantation. Pathol Res Pract 2003; 199(8):565–570. 109. Wright JJ, O’Driscoll G. Treatment of parainfluenza virus 3 pneumonia in a cardiac transplant recipient with intravenous ribavirin and methylprednisolone. J Heart Lung Transplant 2005; 24(3):343–346. 110. Kumar D, Humar A. Emerging viral infections in transplant recipients. Curr Opin Infect Dis 2005; 18(4):337–341. 111. Kumar D, Erdman D, Keshavjee S, et al. Clinical impact of community-acquired respiratory viruses on bronchiolitis obliterans after lung transplant. Am J Transplant 2005; 5(8):2031–2036. 112. Barton TD, Blumberg EA. Viral pneumonias other than cytomegalovirus in transplant recipients. Clin Chest Med 2005; 26(4):707–720, viii. 113. Slifkin M, Doron S, Snydman DR. Viral prophylaxis in organ transplant patients. Drugs 2004; 64(24):2763–2792. 114. Mazzone PJ, Mossad SB, Mawhorter SD, Mehta AC, Mauer JR. Cell-mediated immune response to influenza vaccination in lung transplant recipients. J Heart Lung Transplant 2004; 23(10):1175–1181. 115. Vilchez RA, McCurry K, Dauber J, et al. Influenza virus infection in adult solid organ transplant recipients. Am J Transplant 2002; 2(3):287–291. 116. Ison MG, Hayden FG. Viral infections in immunocompromised patients: what’s new with respiratory viruses? Curr Opin Infect Dis 2002; 15(4):355–367. 117. Garbino J, Gerbase MW, Wunderli W, et al. Lower respiratory viral illnesses: improved diagnosis by molecular methods and clinical impact. Am J Respir Crit Care Med 2004; 170(11):1197–1203 (Epub 2004 Sep 10). 118. Haydar AA, Denton M, West A, Rees J, Goldsmith DJ. Sirolimus-induced pneumonitis: three cases and a review of the literature. Am J Transplant 2004; 4(1):137–139. 119. Singh N. Posttransplant fever in critically ill transplant recipients. Infectious Complications in transplant patients Kluwer Academic publishers 2000; N. Singh, JM Aguado editors. ISBN 0-7923-7972-1:113-132. 120. Ho MC, Wu YM, Hu RH, et al. Surgical complications and outcome of living related liver transplantation. Transplant Proc 2004; 36(8):2249–2251. 121. Rerknimitr R, Sherman S, Fogel EL, et al. Biliary tract complications after orthotopic liver transplantation with choledochocholedochostomy anastomosis: endoscopic findings and results of therapy. Gastrointest Endosc 2002; 55(2):224–231. 122. Piecuch J, Witkowski K. Biliary tract complications following 52 consecutive orthotopic liver transplants. Ann Transplant 2001; 6(1):36–38. 123. Testa G, Malago M, Broelseh CE. Complications of biliary tract in liver transplantation. World J Surg 2001; 25(10):1296–1299. 124. Elola-Olaso AM, Diaz JC, Gonzalez EM, et al. Preliminary study of choledochocholedochostomy without T tube in liver transplantation: a comparative study. Transplant Proc 2005; 37(9):3922–3923. 125. Stange